Dwelling - Pedro P. Ferreira · the sphere, conjuring up a transparent lifeworld which is perceived...

Part II

Dwelling

INTRODUCTION

The chapters in this part explore various aspects of what I have called the dwelling perspec-tive. By this I mean a perspective that treats the immersion of the organism-person in anenvironment or lifeworld as an inescapable condition of existence. From this perspective,the world continually comes into being around the inhabitant, and its manifoldconstituents take on significance through their incorporation into a regular pattern of lifeactivity. It has been rather more usual, in social and cultural anthropology, to supposethat people inhabit a world – of culture or society – to which form and meaning havealready been attached. It is assumed, in other words, that they must perforce ‘construct’the world, in consciousness, before they can act in it. I refer to this view as the buildingperspective. Each chapter explores some aspect of the contrast between the building anddwelling perspectives, in relation to such topics as the significance of architecture, theperception of the landscape, the idea of environmental change, the practice of wayfinding,and the properties of vision and hearing. In order to lay a foundation for these explo-rations, however, I begin in Chapter Nine with a general introduction to anthropologicaltheories of perception and cognition. The fundamental question that all such theories seekto address is the following: why should people from different cultural backgrounds perceivethe world in different ways?

In the first part of the chapter I outline the history of anthropological attempts toanswer this question, starting with the classical work of Emile Durkheim, through influ-ential statements by Edmund Leach, Clifford Geertz and Mary Douglas, to the morerecent development of the field known as cognitive anthropology. Throughout this history,the assumption has persisted that people construct the world, or what for them is ‘reality’,by organising the data of sensory perception in terms of received and culturally specificconceptual schemata. But in recent anthropology, this assumption has been challenged byadvocates of ‘practice theory’, who argue that cultural knowledge, rather than beingimported into the settings of practical activity, is constituted within these settings throughthe development of specific dispositions and sensibilities that lead people to orient them-selves in relation to their environment and to attend to its features in the particular waysthat they do. In the second part of Chapter Nine, I assess the relevance for anthropo-logical understanding of alternative approaches drawn from cognitive science, ecologicalpsychology and phenomenology. Though my conclusion is that anthropology has moreto gain from an alliance with ecological psychology than with cognitive science, and thatsuch an alliance accords well with a phenomenology of dwelling, there are still problemsto be faced in overcoming the dichotomy between culture and biology, in reconciling a

121113451116789101211131114567892012345678930123456789401234567118

phenomenology of the body with an ecology of mind, and in translating the overall theo-retical perspective into a practicable programme of research.

Chapter Ten explores how a dwelling perspective might affect our understanding of thesimilarities and differences between the ways in which human beings and other animalscreate environments for themselves. I am concerned, in particular, with the meaning ofarchitecture, or that part of the environment conventionally described as ‘built’. I start bydocumenting the transition in my own thinking from a ‘building perspective’, accordingto which worlds are made before they are lived in, to a ‘dwelling perspective’, accordingto which the forms people build, whether in the imagination or on the ground, only arisewithin the current of their life activities. Drawing on Jakob von Uexküll’s notion ofUmwelt, I show how we might distinguish between human and non-human constructionsin the terms of the building perspective, on the basis of the presence or absence of anintentional project of design. This argument, however, implies the existence of some kindof threshold in human evolution, beyond which our ancestors were able to author theirown projects. This idea has motivated the search for a point of origin for humanity ingeneral, and for human architecture in particular. Through the adoption of a dwellingperspective, influenced by the philosophy of Martin Heidegger, I show that the point oforigin is illusory. There can, then, be no absolute distinction between ‘natural’ and ‘arti-ficial’ structures. Buildings, like other environmental structures, are never complete butcontinually under construction, and have life-histories of involvement with both theirhuman and non-human inhabitants. Whether, at a certain point in its life history, a struc-ture looks to us like a building or not will depend on the extent and nature of humaninvolvement in its formation.

In Chapter Eleven I turn to what I consider to be the unifying themes of archaeologyand sociocultural anthropology: namely, landscape and temporality. This chapter is anattempt to show how the temporality of the landscape might be understood by way of adwelling perspective. I first set out to clarify the meaning of ‘landscape’ by contrast tothe concepts of land, nature, space and environment. I then introduce the notion of‘taskscape’ to denote a pattern of dwelling activities. The intrinsic temporality of thetaskscape, I argue, lies in its rhythmic interrelations or patterns of resonance. At first glancethe opposition between landscape and taskscape seems to mirror that, in the field of art,between painting and music. However by considering how taskscape relates to landscape,the distinction between them is ultimately dissolved, and the landscape itself is shown tobe fundamentally temporal. I illustrate the thesis of the temporality of the landscapethrough an analysis of the scene depicted by Pieter Bruegel the Elder in his painting Theharvesters. In conclusion, I criticise the view that a properly cultural ecology would be onethat would go beyond strictly pragmatic concerns with the conditions of adaptation tofocus on the multiple layers of symbolic meaning with which people cover over their envi-ronments. For meaning, I contend, does not cover the world but is immanent in thecontexts of people’s pragmatic engagements with its constituents. But the discovery ofmeaning in the landscape has to begin from a recognition of its temporality, and in thislies the essence of archaeological investigation.

The significance of the contrast between building and dwelling perspectives for cosmo-logical conceptions of ‘the earth’ is my theme in Chapter Twelve. I argue that the imageof the earth as a globe, implied in such phrases as ‘global environmental change’, is onethat actually expels humanity from the lifeworld, such that rather than the environmentsurrounding us, it is we who have surrounded it. Far from reintegrating human societyinto the world of nature, the idea of the earth as a solid globe of opaque materiality marks

123456789101234567892012345678930123456789401234567118

Dwelling• 154 •

their final separation. Thus the biodiversity of locally distributed life-forms presents itselfto a universal, globally distributed humanity. The conservation ethic entailed in such aglobal vision, which places nature on the inside and humanity on the outside, is at onceecocentric and anthropocircumferential. Against this, I examine the contrasting image ofthe sphere, conjuring up a transparent lifeworld which is perceived by its inhabitants fromwithin. This image, which is characteristic of the cosmologies of pre-modern societies, isgenuinely anthropocentric, but in a way that counterposes neither humanity and nature,nor the local and the global. I show how the shift from a spherical to a global perspec-tive marks the triumph of technology over cosmology. But it also leads to the systematicdisempowerment of local communities, taking from them – in the name of preservingbiodiversity – the responsibility to care for their own environments.

From my discussion of the landscape and of the topological image of the globe inChapters Eleven and Twelve, it is clear that in the building perspective (as in the genealog-ical model of Chapter Eight) the earth is presented to humanity as a surface to be occupiedrather than a world to be inhabited. It is further supposed that the disposition of thingsand places on this surface is known by representing it, either in the mind or on paper,in the form of a map. Thus to know where one is entails identifying one’s current posi-tion with a corresponding location on the map, and to find one’s way from one positionto another is to navigate by means of it. In Chapter Thirteen I take a critical look at thenotion of the map, and its application in anthropological studies of wayfinding and navi-gation. I argue that while dwelling in the world entails movement, this movement is notbetween locations in space but between places in a network of coming and going that Icall a region. To know one’s whereabouts is thus to be able to connect one’s latest move-ments to narratives of journeys previously made, by oneself and others. In wayfinding,people do not traverse the surface of a world whose layout is fixed in advance – as repre-sented on the cartographic map. Rather, they ‘feel their way’ through a world that is itselfin motion, continually coming into being through the combined action of human andnon-human agencies. I develop a notion of mapping as the narrative re-enactment of jour-neys made, and of maps as the inscriptions to which such re-enactments may possiblygive rise. However, the building perspective enshrined in modern science splits mappinginto the phases of mapmaking and map-using, and likewise splits wayfinding into the twinprojects of cartography and navigation.

In Chapter Fourteen I turn to a problem in the anthropology of the senses. Does abuilding perspective imply the hegemony of vision? Is hearing the predominant sense ofdwelling? To regain an appreciation of human dwelling in the world is it necessary torebalance the sensorium, giving greater weight to the ear, and less to the eye, in the ratioof the senses? Many philosophers and historians have noted the ‘ocularcentrism’ of theWestern tradition, its privileging of sight over the other senses as a source of human know-ledge. Anthropologists, for their part, have stressed the importance of hearing in thesensorium of many non-Western peoples. Yet the comparison remains couched in termsof a dichotomy between vision and hearing whose roots lie firmly in the intellectual historyof the West. In the terms of this dichotomy, vision is distancing, objectifying, analytic,and atomising; hearing is unifying, subjective, synthetic and holistic. Vision represents anexternal world of being, hearing participates in the inwardness of the world’s becoming:the former is inherently static, the latter suspended in movement. Whereas one hearssound, one does not see light, but only the things off whose surfaces light is reflected.This is why hearing is supposed to penetrate the inner, subjective domain of thought andfeeling in a way that vision cannot. It is also why Western thought, for all its dependence

123456789101231114567892012345678930123456789401234567118

Introduction to Part II • 155 •

on the written word, and in apparent contradiction to its elevation of sight as the ‘noblest’of the senses, has tended to treat writing (which is seen) as inferior to speech (which isheard).

But ethnography suggests that people in non-Western societies do not regard visionand hearing as radically opposed, but rather as virtually interchangeable. Nor does theirapparent emphasis on understanding through sensory participation rather than externalobservation entail a bias towards hearing over vision. For many, vision remains paramount.But it is a vision that is non-representational, a matter of watching rather than seeing.Like hearing, it is caught in the flow of time and bodily movement. One can, in short,dwell just as fully in the world of visual as in that of aural experience: indeed for themost part these worlds are one and the same. That this point has been missed in theanthropology of the senses is due to its tendency to treat sensory experience as but avehicle for the expression of extra-sensory, cultural values. The key question, I conclude,is: what is the relationship between the cultural evaluation of the senses and the ways inwhich they are practically deployed in acts of perception?

123456789101234567892012345678930123456789401234567118

Dwelling• 156 •

Chapter Nine

Culture, perception and cognition

There is one question that, perhaps more than any other, motivates anthropological inquiry.Take people from different backgrounds and place them in the same situation; they arelikely to differ in what they make of it. Indeed such difference is something that everyanthropologist experiences in the initial phases of fieldwork. But why should this be so?How do we account for it? In their attempts to answer this question, anthropologists havecome up against some of the most contested issues in the psychology of perception andcognition. My purpose in this chapter is to show how they have dealt with these issues.The chapter is divided into two parts. In the first part I trace something of the historyof the problem over the past century of anthropological thought. In the second, I go onto assess the relevance for anthropological understanding of alternative approaches drawnfrom cognitive science, ecological psychology and phenomenology. This is a considerableagenda, and in the space of a single chapter I can do no more than touch on the manyquestions raised.

I

SOCIAL ANTHROPOLOGY

In British social anthropology (as distinct from American cultural anthropology) thinkingabout perception and cognition goes back to the classical work of Emile Durkheim, himselfone of the founding fathers of what was then the new science of sociology. In his mani-festo for the new discipline, The rules of sociological method (first published in 1895),Durkheim adamantly opposed all attempts to explain social phenomena in terms of thepsychological properties of individuals. As he famously declared, ‘every time a socialphenomenon is directly explained by a psychological phenomenon, we may rest assuredthat the explanation is false’ (1982[1895]: 129). If sociology is a kind of psychology,Durkheim thought, its object of study must be the mind of society, not of the individual.This mind, the consciousness of the collectivity, was supposed to have emergent proper-ties of its own, in no way reducible to the given properties of individuals as inscribed inhuman nature. But it was not until the concluding chapter of his greatest work, Theelementary forms of the religious life, that Durkheim explicitly spelled out the relationbetween the consciousness of the individual and that of the collectivity – ‘the highest formof the psychic life’ (1976[1915]: 444). He did so in terms of a thoroughgoing distinc-tion between sensation and representation.

The distinction was made on two grounds. The first lies in the contrast between theephemerality of sensations and the durability of representations. Every sensation, Durkheim

123456789101231114567892012345678930123456789401234567118

Culture, nature, environment • 157 •

argued, is tied to a particular moment that will never recur, for even if – at a subsequentpoint in time – the thing perceived has not changed, the perceiver will no longer be thesame. We are nevertheless able to represent our experience, and so to know what we haveperceived, by catching perceptual images that would otherwise float by on the stream ofconsciousness within the mesh of a system of concepts that remains somehow aloof fromthis sensory agitation (in a ‘different portion of the mind’, Durkheim suggested, that ismore calm and serene). Like language, which is the medium in which concepts areexpressed (‘for every word translates a concept’), the conceptual system has a kind ofstability: it endures, whilst the stream of consciousness flows on (Durkheim 1976[1915]:433).

Secondly, whereas sensations are private and individual, representations are public andsocial. Since sensations consist in the reactions of the organism to particular externalstimuli, there is no way in which a sensation can be made to pass directly from one indi-vidual consciousness to another. If people are to share their experiences they must talkabout them, and to do that these experiences must be represented by means of concepts,which in turn may be expressed in words whose meanings are established within a commu-nity of speakers by verbal convention. Thus collective representations serve as a kind ofbridge between individual consciousnesses that are otherwise closed to each other,furnishing them with a means of mutual understanding. ‘The concept is an essentiallyimpersonal representation; it is through it that human intelligences communicate’(Durkheim 1976[1915]: 433–4).

Following Durkheim’s lead, British social anthropologists carried on with the compar-ative study of collective representations – otherwise known as ‘social structures’ – withoutpaying much attention to the psychological premises on which such study rested. Fiftyyears later, two of the most influential social anthropologists of the day, Edmund Leachand Mary Douglas, could still pose the problem of perception and cognition in very muchthe same terms. Given that the world of our immediate, sensory experience is a formlessand continuous flux in which nothing is the same from one moment to the next, howcan we know what we perceive? To recognise specific objects and events in the externalworld, Leach claimed, the flux has to be cut up into bounded chunks: thus thought frag-ments the continuum of life as it is lived, and the diversity of culture lies precisely in themanifold ways in which the continuum can be cut. Leach’s first explicit statement of thistheory of perception and cognition was presented in an article on ‘Anthropological aspectsof language’, published in 1964. Here he argued that the categories of language providethe ‘discriminating grid’ which, laid over the continuous substrate of raw experience,enables the speaker to tell one thing from another, and so to see the world ‘as beingcomposed of a large number of separate things, each labelled with a name’ (1964: 34).As the child learns its mother-tongue, thereby taking on board a conventional system ofnamed categories, so its environment literally takes shape before its very eyes.

Two years later, Mary Douglas published her seminal study, Purity and danger. Here,too, we find the same basic idea: that in perception the world is constructed to a certainorder, through the imposition of culturally transmitted form upon the flux of experience.

As perceivers we select from all the stimuli falling on our senses only those that interestus, and our interests are governed by a pattern-making tendency . . . In a chaos ofshifting impressions, each of us constructs a world in which objects have recognisableshapes, are located in depth, and have permanence.

(1966: 36)

123456789101234567892012345678930123456789401234567118

Dwelling• 158 •

As with Leach, the roots of Douglas’s thinking lie in Durkheim’s theory of knowledge.This theory, as we have seen, effectively divides the human subject into two mutuallyexclusive parts. One part, fully immersed in the sensate, physical world, is continuallybombarded by stimuli which are registered in consciousness as a ‘chaos of shifting impres-sions’. The other part, however, stands aside from this engagement, and is untouched byit. Here are located the conceptual categories that sort the sensory input, discarding orsuppressing some elements of it while fitting the remainder into a pre-existing, sociallyapproved schema. Crucially, then, perception is a two-stage phenomenon: the first involvesthe receipt, by the individual human organism, of ephemeral and meaningless sense data;the second consists in the organisation of these data into collectively held and enduringrepresentations.

CULTURAL ANTHROPOLOGY

The rigid distinction between social and psychological phenomena that British socialanthropology took from Durkheim was not matched by the parallel, North Americantradition of cultural anthropology. The founder of this latter tradition, Franz Boas, consis-tently adopted the position that the patterned integration of culture, as a system of habits,beliefs and dispositions, is achieved on the level of the individual rather than having itssource in some overarching collectivity, and is therefore essentially psychological in nature.Accordingly, American cultural anthropologists of the mid-twentieth century paid a greatdeal of attention to the way in which the individual personality is fashioned out of thecultural materials available to it. In two respects, however, subsequent developments ledto the establishment of a view of perception and cognition more closely in line with thatespoused by British writers. The first lay in the separation of culture, as a body of trans-missible knowledge, from patterns of observable behaviour. Already in the writings ofClyde Kluckhohn, and in the review of concepts of culture that Kluckhohn compiled incollaboration with Alfred Kroeber, we find a stress on culture as an internalised systemof rules and meanings as distinct from manifest behaviour patterns and their artefactualproducts (Kluckhohn 1949: 32, Kroeber and Kluckhohn 1952: 114). And in 1957, WardGoodenough confirmed this separation in his much cited definition of culture as ‘what-ever it is one has to know or believe in order to operate in a manner acceptable to [asociety’s] members’ (cited in D’Andrade 1984: 89).

The distinction between culture and behaviour was once again reiterated, this time byClifford Geertz, in an influential article first published in 1966, on ‘The impact of theconcept of culture on the concept of man’. Culture, Geertz argued, ‘is best seen not ascomplexes of concrete behavior patterns – customs, usages, traditions, habit clusters – . . . but as a set of control mechanisms – plans, recipes, rules, instructions (what computerengineers call “programs”) – for the governing of behavior’ (Geertz 1973: 44). He never-theless took strong exception to the view, attributed to Goodenough, that the place tofind these control mechanisms is inside the heads of individuals.1 Herein, then, lay thesecond development: having split culture from behaviour, the former was removed fromthe minds of individuals and reinscribed on the level of the collectivity. In a move redo-lent of Durkheim’s earlier formulation, Geertz insisted that the domain of cultural symbolsis social rather than psychological, public rather than private. Their natural place of abodeis in the intersubjective space of social interaction – ‘the house yard, the marketplace, andthe town square’ – whence they are ‘used to impose meaning upon experience’ (1973:44–5). For any one individual, the range of symbolic meanings which can be drawn upon

123456789101231114567892012345678930123456789401234567118

Culture, perception and cognition • 159 •

is more or less given by what is current in the community into which he or she is born.But without the guidance provided by significant symbols, human beings would be hope-lessly lost, unable to establish their bearings in the world. For unlike other creatures whoseactivities are more closely controlled by innate response mechanisms, humans depend ona substantial input of additional information, learned rather than innate, in order to func-tion adequately in their normal environments. ‘Undirected by culture patterns – organizedsystems of significant symbols – man’s behavior would be virtually ungovernable, a merechaos of pointless acts and exploding emotions, his experience virtually shapeless’ (Geertz1973: 46).

Despite his different intellectual roots, in American cultural anthropology rather thanBritish social anthropology, Geertz came to conclusions remarkably similar to those thatwere being drawn at the same time by Douglas, and that I have already touched upon.Both Geertz and Douglas took culture to comprise a framework of symbolic meanings,common to a community and relatively impervious to the passage of time and genera-tions, which gives shape to the raw material of experience and direction to human feelingand action. Thus to return to our original question: if two individuals from different back-grounds, placed in the same environment, construe it in different ways, the reason wouldbe that each has brought a different symbolic system to bear in organising the samematerial of sensory experience. Granted, then, that every community has its own partic-ular system for the organisation of experience, anthropological attention naturally cameto focus on cultural variation in the organisational principles involved. Geertz, as we haveseen, claimed that such principles were to be found in the publicly accessible space ofsocial discourse, and not in the interiority of the mind. But others, taking their cue moredirectly from Goodenough, insisted that cultural cognition can only take place by way ofshared conceptual schemata lodged in the minds of individuals. Their aim was to uncoverthese schemata, and it gave rise, in the late 1960s, to a field of inquiry known rathergenerally as ‘cognitive anthropology’, though in a narrower and more restricted form as‘ethnoscience’ (Tyler 1969).

COGNITIVE ANTHROPOLOGY

The problem for the cognitive anthropologist, Tyler explains, ‘is to discover how otherpeople create order out of what appears to him to be utter chaos’ (1969: 6). They do so,it is supposed, by grouping the infinitely variable phenomena of the experienced worldinto a finite set of named, hierarchically ordered classes. This is done by attending onlyto those perceptual cues that differentiate things as belonging to one class rather thananother, while ignoring those that would indicate the uniqueness of every member of aclass. But the ordering principles that govern this process of selective attention are givenin the mind, not in the world. ‘There is nothing’, Tyler asserts, ‘in the external worldwhich demands that certain things go together and others do not’ (1969: 7). In otherwords, the principles of classification are arbitrary and subjective with regard to the worldwhose phenomena are to be classified. They are to be discovered through the formalanalysis of responses provided by native informants to a series of questions of the form‘is this thing here a kind of X?’, ‘what other kinds of X are there?’, ‘is X a kind of Y?’,and so on, all of which are designed by the investigator to elicit precisely the distinctionshe or she is looking for.

Despite early promise, the project of cognitive anthropology soon ran into difficulties.An enormous amount of effort was put into mapping out rather limited semantic domains

123456789101234567892012345678930123456789401234567118

Dwelling• 160 •

– for example of kinship terms, plant and animal taxonomies or colour classifications –without bringing any comparable advance in understanding how people actually negotiatetheir relationships with one another, and with their non-human environments, in the usualcourse of everyday life. It became apparent that the key to such negotiation lay in a certainflexibility in the use of concepts and a sensitivity to context that was disregarded by formalsemantic analysis. The neatly ordered paradigms and taxonomies yielded by this methodof analysis seemed to be artefacts of anthropologists’ techniques of controlled elicitationrather than having any counterpart in the cognitive organisation of the people studied.The specialised tasks of naming and discrimination that the latter were expected to performwere not, after all, ones that they would have ordinarily encountered. Indeed the abilityto name things correctly is but a small and relatively insignificant part of what a personneeds to know in order to get by in the world, so that the greater part of cultural know-ledge had still to be uncovered. Above all, cognitive anthropology was unable to grasp thesource of human motives: one learned no more from an analysis, say, of kinship termi-nology about people’s feelings for one another than one might learn from the grammarof a language about why its speakers say the things they do.

In recent years, and partly in response to these objections, cognitive anthropology hasresurfaced in a new guise, as the investigation of what are now called ‘cultural models’.Introducing a seminal volume of essays on Cultural models in language and thought, NaomiQuinn and Dorothy Holland define such models as ‘presupposed, taken-for-granted modelsof the world that are widely shared . . . by the members of a society and that play anenormous role in their understanding of that world and their behaviour in it’ (1987: 4).They differ from the classificatory schemas identified by earlier cognitive anthropologistsin three major ways. First, rather than dividing up the continuum of experience intonamed categories, cultural models offer a description of the world framed in terms ofnetworks of interconnected images or propositions, in which objects, events and situationstake on regular, prototypical forms. Actual experience in the real world is then organisedby matching it to the prototypical scenarios built into the simplified worlds of the culturalmodels, and these, in turn, furnish conventional guidelines for action. Secondly, althoughlinguistic data provide important clues to underlying cultural knowledge, it cannot beassumed that word meanings stand to components of the cultural model in a simple rela-tion of one-to-one correspondence. The relation is rather complex and indirect, and canonly be grasped through an analysis of the richly textured material of ordinary discourse.Thirdly, cultural models – to the extent that they are fully internalised – do not merelydescribe or represent the world, they also shape people’s feelings and desires. That is tosay, they can have ‘motivational force’ (D’Andrade 1992: 28). As Claudia Strauss argues,in her introduction to a recent volume dedicated to the demonstration of this point, therealm of cognition is inseparable from the realm of affect; thus cultural models should beunderstood as ‘learned, internalised patterns of thought-feeling’ (Strauss 1992: 3).

Despite these fairly radical revisions, the programme of cognitive anthropology remainsbasically unchanged. Starting from the premise that culture consists in a corpus of inter-generationally transmissible knowledge, as distinct from the ways in which it is put to usein practical contexts of perception and action, the objective is to discover how this know-ledge is organised. Moreover the assumptions on which the programme rests are much asthey were in Durkheim’s day. They are that cognition consists of a process of matchingsensory experience to stable conceptual schemata, that much if not all of the order thatpeople claim to perceive in the world – and especially the social world – is imposed bythe mind rather than given in experience, that people are able to understand one another

123456789101231114567892012345678930123456789401234567118


to the extent that their cultural orderings are founded on consensus (such that the limitsof consensus define the boundaries of society), and that the acquisition of such orderingsinvolves a process of internalisation. These assumptions have not, however, gone unchal-lenged – indeed there is a powerful movement within contemporary anthropology thatwould reject them altogether. One of the most influential figures in this movement hasbeen Pierre Bourdieu, who in a series of works has attempted to show how cultural know-ledge, rather than being imported by the mind into contexts of experience, is itselfgenerated within these contexts in the course of people’s involvement with others in thepractical business of life. Through such involvement, people acquire the specific disposi-tions and sensibilities that lead them to orient themselves in relation to their environmentand to attend to its features in the particular ways that they do. These dispositions andsensibilities add up to what Bourdieu calls the habitus (1990: 52–65).2

THE THEORY OF PRACTICE

Like the ‘cultural model’ of cognitive anthropology, the habitus of Bourdieu’s theory of prac-tice could be described as a pattern of thought-feeling. The similarity ends there, however.For thinking and feeling, in Bourdieu’s account, do not go on in an interior subjective (or intersubjective) space of images and representations but in the space of people’s actualengagement in the settings of practical activity. Whereas cultural models are supposed toexist independently of, and prior to, their application in particular situations of use – suchas in doing things or making things, or in the interpretation of experience – the habitusexists only as it is instantiated in the activity itself. In other words, the habitus is notexpressed in practice, it rather subsists in it.3 What Bourdieu has in mind is the kind of practical mastery that we associate with skill – a mastery that we carry in our bodies andthat is refractory to formulation in terms of any system of mental rules and representations.Such skill is acquired not through formal instruction, but by routinely carrying out specifictasks involving characteristic postures and gestures, or what Bourdieu calls a particular bodyhexis. ‘A way of walking, a tilt of the head, facial expressions, ways of sitting and of usingimplements’ – all of these, and more, comprise what it takes to be an accomplished prac-titioner, and together they furnish a person with his or her bearings in the world (Bourdieu1977: 87). And if people from different backgrounds orient themselves in different ways,this is not because they are interpreting the same sensory experience in terms of alternativecultural models or cognitive schemata, but because, due to their previous bodily training,their senses are differentially attuned to the environment.

In the anthropological study of cognition this kind of approach is perhaps best repre-sented in the work of Jean Lave. Her book Cognition in Practice (1988) is a manifestofor an ‘outdoor psychology’ – that is, a psychology that would take as its unit of analysis‘the whole person in action, acting within the settings of that activity’ (1988: 17).Cognition, in Lave’s view, is not a process that goes on ‘inside the head’, whose productsare representations that bear some complex relation to the world outside, but rather asocial activity that is situated in the nexus of ongoing relations between persons and theworld, and that plays its part in their mutual constitution. It is a process wherein bothpersons, as knowledgeable social agents, and the settings in which they act, continuallycome into being, each in relation to the other. Thus thinking is inseparable from doing,thought is ‘embodied and enacted’, and cognition is ‘seamlessly distributed across persons,activity and setting’ (1988: 171). To study cognition is to focus on the modus operandinot of the mind, in organising the bodily data of sense, but of the whole body-person in

123456789101234567892012345678930123456789401234567118

Dwelling• 162 •

the business of dwelling in the world. And if knowledge is shared it is because peoplework together, through their joint immersion in the settings of activity, in the process ofits formation.

What, then, becomes of the models and schemata of the cognitive anthropologists? Arethey merely artefacts of analytic abstraction, products of attempts by anthropologicalobservers to represent manifest behaviour as the output of formal programmes? Or dothey, to the contrary, offer clues to basic truths about the way the human mind works?The answers to these questions hinge on more fundamental differences of approach whichdivide psychologists as much as anthropologists. Roughly speaking, the division is betweenadvocates of cognitive science on the one hand, and their critics on the other, who findinspiration in an ecological or phenomenological perspective on perception and cognition.These differences of approach, and some of their implications for anthropology, arereviewed in the next part of this chapter.

II

COGNITIVE SCIENCE

In the field of psychology, cognitive science emerged as an alternative to behaviourism inthe 1950s, alongside the development of the digital computer. Its founding axioms arethat people come to know what is ‘out there’ in the world by representing it in the mind,in the form of ‘mental models’, and that such representations are the result of a compu-tational process working upon information received by the senses. The functioning of themind, then, can be compared to the operation of a computer program, and the relationbetween mind and brain to that between the program and the ‘hardware’ in which it isinstalled (Johnson-Laird 1988). But the computing analogy also found its way into cogni-tive anthropology – I have already referred to Geertz’s (1973: 44) likening of culturalcontrol mechanisms to computer software – where it was similarly supposed that the mindis equipped with programmes that construct internal representations of the environmentfrom the data of sensation, and deliver appropriate plans for action (D’Andrade 1984:88–9). Whereas cognitive scientists, however, have by and large been concerned to discoveruniversals of human cognition, which are attributed to innate structures established in thecourse of evolution under natural selection, cognitive anthropologists have sought toaccount for human perception and action in terms of acquired schemata or programmesthat differ from one culture to another.

How, then, should we view the relation between these two projects? Are they contra-dictory or mutually compatible? D’Andrade (1981: 181–2) tackles this issue by consideringthe fit between programmes and processors. By programmes he means the informationalcontent of transmitted culture – what is ‘passed along’ from generation to generation. Byprocessors he means the apparatus of acquisition that makes such transmission possible,an apparatus that is assumed to be common to all human minds. According to this division, cognitive anthropology is concerned with the diversity of cultural content, andwith the way in which its organisation is constrained by invariant properties of theprocessing devices that govern its acquisition, while cognitive psychology is concerned withthe structure and functioning of the devices themselves, and the way in which they workon all kinds of information (including cultural information). This formulation, however,begs a critical question. Granted that mental representations are the products of a

123456789101231114567892012345678930123456789401234567118


processing of information by acquired cultural programmes, what is the source of theprocessing apparatus of which these programmes are themselves products? This apparatus,it seems, must already be in place prior to the acquisition of culture; hence its design andoperation must be innately specified. In short, the theory that all human cognition isgrounded in culturally specific schemata can hold only on condition that human beingscome universally pre-equipped with the structures necessary to enable these schemata tobe acquired in the first place.

This is precisely the conclusion reached by Dan Sperber (1985), in the context of hiscritique of cultural relativism – the doctrine, long ascendant in anthropology, that peoplein different cultures inhabit different cognitive (or rather, cognisable) worlds, each withits own criteria of rationality and judgement. Relativists argue that just as every non-human animal species, depending on its evolved cognitive organisation, can only knowthe world in its own particular way, so also every human culture is locked into the cogni-tive framework of a unique worldview. But whereas species differences supposedly have agenetic basis, cultural differences are assumed to be entirely independent of geneticconstraint. Thus cultural relativists tend to imagine that theirs is a position opposed toan innatist view of the human mind, and that evidence for the diversity of incommen-surate worldviews only goes to prove that the underlying structures of human cognitionare genetically underdetermined and malleable to the effects of experience.

Yet in this, Sperber shows, they are mistaken. Relativists, he contends, have failed toattend to the psychological implications of their assumption that human behaviour isrooted in tradition rather than heredity. Had they done so, they would have realised thata creature capable of taking on not just one form of life but any one of a very large numberof possible alternative forms would require more rather than less by way of innate program-ming. On the basis of a formal logical argument, Sperber concludes that ‘the greater thediversity of the cultures that humans are capable of acquiring, the greater the complexityof the innate learning abilities involved’ (1985: 43). Thus the relativists’ appeal to humancultural diversity is not at all contrary to the universalist claims of cognitive science; ratherit depends upon them.

Though the logic of Sperber’s argument may be impeccable, it rests on a foundationthat is far from secure – namely, that cultural knowledge takes the propositional (or semi-propositional) form of beliefs, ‘representations acquired through social communication andaccepted on the ground of social affiliation’ (1985: 59). Underlying the commonsenseunderstanding of the culturally competent actor is supposed to lie a huge database of suchrepresentations, which provide all the information necessary to generate appropriateresponses under any given environmental circumstances. Yet as many critics of cognitivescience have pointed out, and as the failure of attempts to replicate human skills in thedesign of expert systems has amply demonstrated (Dreyfus and Dreyfus 1987), even thesimplest and most routine of everyday tasks are refractory to codification in propositionalform. By and large, these tasks are not represented (save in the notebooks of observers),nor are such representations communicated in learning situations. Most cultural learningtakes place through trial-and-error and practice, albeit in socially structured situations, andalthough beginners may need to follow rules, these rules structure the situation of learningand do not themselves form any part of the content of what is learned. For the skilledpractitioner consults the world, rather than representations (rules, propositions, beliefs)inside his or her head, for guidance on what to do next.4 As Andy Clark puts it, whyshould we go to the trouble of modelling the world when ‘we can use the world as itsown best model’ (Clark 1997: 29–30, see also Chapman 1991: 20)?

123456789101234567892012345678930123456789401234567118

Dwelling• 164 •

Faced with the evident artificiality of depicting cultural knowledge in algorithmisedform as a set of programmes, acquired by means of a processing device that is somehowconstituted in advance of ontogenetic development, cognitive science has come up withan alternative model of the way the mind works. Instead of positing one giant processorwith a massive capacity for information storage and retrieval, it is suggested that the mindconsists of a very large number of small, simple processors, massively interconnected, alloperating in parallel, and receiving inputs and delivering outputs to each other along thecountless pathways linking them. Crucially, a system so constituted can learn from expe-rience, not by taking on new informational content, but by adjustments to the differentialstrengths of the connections among processing units. In other words, knowledge is acquiredthrough the establishment of particular patterns of connection: any processor may there-fore be involved in the representation of diverse experiences; conversely the representationof any experience may be distributed across many processors (Johnson-Laird 1988: 174).This so-called ‘connectionist’ model of the mind has a certain anthropological appeal –thus cognitive anthropologists such as D’Andrade (1990: 98–9) have noted that the prop-erties of cultural models are precisely what would be expected from the operation ofparallel processing networks, while Bloch (1991) has suggested that the acquisition of prac-tical skills may best be understood in terms of the development of tightly connectednetworks dedicated to particular domains of cognition (for a more extended review, seeD’Andrade 1995: 143–9).

Despite its greater realism, connectionism remains open to much the same criticismsthat have been levelled against earlier versions of artificial intelligence (Dreyfus 1992). Forultimately, it is still grounded in the Cartesian ontology that is basic to the entire projectof cognitive science – an ontology that divorces the activity of the mind from that of thebody in the world. Thus the body continues to be regarded as nothing more than aninput device whose role is to receive information to be ‘processed’ by the mind, ratherthan playing any part in cognition itself. And beyond that, the world is supposed to existas a domain of problems to be solved, or as a field for the enactment of solutions reached,rather than as a resource for problem solving (Clark 1997: 83–4). Connectionists, Clarkadmits, ‘inherit a distressing tendency to study disembodied problem solving and to optfor abstract, symbolically defined input-output mappings’ (1997: 80). What they fail torecognise is that the processing loops that yield intelligent action are not confined to someinterior space of mind, confined within the skull, but freely penetrate both the body andits environment. This failure is deeply rooted in the history of twentieth-centurypsychology. It lies, as Edward Reed (1987: 144–5) has shown, in the founding assump-tions of the behaviourist theory that cognitive science claims to have overthrown: namelythat perception is based on discrete bodily sensations touched off by external stimuli, andthat action is based on the corresponding bodily responses.

The objection to behaviourism was that, as a theory, it was incomplete: the simplelinkage of stimulus and response was considered insufficient to account for the know-ledgeability of actors or the productivity of their actions. To complete the picture, cognitivescientists posited a mental processing device that would convert the stimulus input intoknowledge, and generate plans for the delivery of meaningful responses. There is howeveranother way out of behaviourism, and this is to treat the perceiving organism not as apassive recipient of stimuli but as an active agent who purposively seeks out informationthat would specify the meaningful properties of his or her environment. This was the pathtaken by James Gibson in his pioneering studies of visual perception, and in doing so helaid the foundations for an approach, known as ‘ecological psychology’, which is radically

123456789101231114567892012345678930123456789401234567118


opposed, in almost every respect, to the project of cognitive science.

ECOLOGICAL PSYCHOLOGY

The point of departure for ecological psychology is the proposition that perceptual activityconsists not in the operation of the mind upon the bodily data of sense, but in the inten-tional movement of the whole being (indissolubly body and mind) in its environment.The emphasis on movement is critical. Cognitive science assumes a static perceiver whohas nothing to go on but transient patterns of sensory excitation that are, in themselves,quite insufficient to specify the objects and events that gave rise to them. Thus the problemof perception, for the cognitive scientist, is to show how these ephemeral and fragmen-tary sense data are reconstructed, in terms of pre-existing schemata or representations, intoa coherent picture of the world. But for Gibson, sensations do not, as such, constitutethe data for perception (Gibson 1979: 55). Rather, what the perceiver looks for are constan-cies underlying the continuous modulations of the sensory array as one moves from placeto place. In visual perception, for example, we do not see patterns of light but objects inour environment. We do so because, as we move about, the pattern of light reaching theeyes from reflecting surfaces in the environment (that is, the ‘optic array’) undergoes agradual transformation. It is the invariants that underly this transformation, and not themomentary patterns of stimulation themselves, that specify what we see. Indeed it isGibson’s contention that the invariant relations that structure the modulations of an opticarray for a moving observer contain all the information necessary to specify the environ-ment. Perception, then, is a matter of extracting these invariants. The perceiver has noneed to reconstruct the world in the mind if it can be accessed directly in this way.

Certain implications follow. First, if perception entails movement, then it must be amode of action rather than a prerequisite for action. For Gibson, perception is an activeand exploratory process of information pickup; far from working on sensations alreadyreceived, it involves the continual movement, adjustment and reorientation of the receptororgans themselves. What is important, he argues, ‘is the looking, listening, touching andsniffing that goes on when the perceptual systems are at work’ (1982[1976]: 397–8).Secondly, if perception is a mode of action, then what we perceive must be a direct func-tion of how we act. Depending on the kind of activity in which we are engaged, we willbe attuned to picking up particular kinds of information. The knowledge obtained throughdirect perception is thus practical, it is knowledge about what an environment offers forthe pursuance of the action in which the perceiver is currently engaged. In other words,to perceive an object or event is to perceive what it affords. Perhaps the most fundamentalcontribution of Gibson’s approach to perception lay in his insight that the informationpicked up by an agent in the context of practical activity specifies what are called the‘affordances’ of objects and events in the environment (Gibson 1979: 127–43).

Thirdly, the information that is potentially available to an agent is inexhaustible: thereis no limit to what can be perceived. Throughout life one can keep on seeing new thingsin an otherwise permanent world, not by constructing the same sense data according tonovel conceptual schemata, but by a sensitisation or ‘fine-tuning’ of the perceptual systemto new kinds of information. Novel perceptions arise from creative acts of discovery ratherthan imagining, and the information on which they are based is available to anyone attunedto pick it up. Finally, and following from the above, one learns to perceive in the mannerappropriate to a culture, not by acquiring programmes or conceptual schemata for organ-ising sensory data into higher-order representations, but by ‘hands-on’ training in everyday

123456789101234567892012345678930123456789401234567118

Dwelling• 166 •

tasks whose successful fulfilment requires a practised ability to notice and to respondfluently to salient aspects of the environment. In short, learning is not a transmission ofinformation but – in Gibson’s (1979: 254) words – an ‘education of attention’. As such,it is inseparable from a person’s life in the world, and indeed continues for as long as heor she lives.

There are clear parallels between the ecological critique, in the field of psychology, ofcognitive science and the critique by practice theorists of cognitive anthropology, whichI reviewed in the first part of this chapter. Both Gibson’s ecological psychology andBourdieu’s theory of practice set out to re-embed perception and cognition within thepractical contexts of people’s ongoing engagement with their environments in the ordi-nary course of life. And both seek to escape from the sterile Cartesian dualisms of mindand nature, subject and object, intellection and sensation, and so on. Yet while the impactof Bourdieu’s work in social and cultural anthropology has been immense, the relevanceof Gibsonian ecological psychology to anthropological theory has been little explored. Anobvious reason for the discrepancy lies in the fact that Gibson himself devoted scant atten-tion to the specifically social and cultural dimensions of human life, preferring – if anything– to downplay the significance of the distinction between human beings and other animals.In developing his theory of affordances, Gibson did devote a brief section to ‘other personsand animals’ in the environment of the perceiver, noting that they have the peculiarcapacity to ‘act back’ or, literally, to interact with the perceiver. Thus ‘behavior affordsbehavior, and the whole subject matter of psychology and of the social sciences can bethought of as an elaboration of this basic fact’ (Gibson 1979: 135). But beyond suggestingthat the perception of mutual affordances in social life involves the same principles ofinformation pickup as are involved in the perception of inanimate objects, Gibson didnot pursue further the implications of this rather sweeping statement.

A recent attempt to develop this neglected aspect of the Gibsonian programme has beenmade by Edward Reed (1988a). The crux of his argument is that social agents can notonly directly perceive their mutual affordances for one another, but also share their directperception of other constituents of the environment. Attuned through prior training andexperience to attending to similar invariants, and moving in the same environment in thepursuit of joint activities, they will pick up the same information (Reed 1988a: 119–20,see Gibson 1982[1967]: 412). Thus, contrary to the axioms of cognitive anthropology,the communion of experience that lies at the heart of sociality does not depend upon theorganisation of sensory data, initially private to each perceiver, in terms of an objectivesystem of collective representations. Rather, sociality is given from the start, prior to theobjectification of experience in cultural categories, in the direct, perceptual involvementof fellow participants in a shared environment (Ingold 1993a: 222–3). This, indeed, iswhat makes anthropological fieldwork possible, for it allows the fieldworker and localpeople to inhabit a common ground of experience, even though each may bring to beara radically different conceptual frame to the task of its interpretation. As Michael Jacksonnotes, ‘by using one’s body in the same way as others in the same environment one findsoneself informed by an understanding which may then be interpreted according to one’sown custom or bent, yet which remains grounded in a field of practical activity andthereby remains consonant with the experience of those among whom one has lived’ (1989:135).

The environment of joint practical activity should not, however, be confused with thephysical world of ‘nature’ (Gibson 1979: 8). For the world can appear in this latter guiseonly to a creature that can disengage itself – or imagine itself to be disengaged – from

123456789101231114567892012345678930123456789401234567118


the processes of its own material life. But the world we inhabit does not confront us, it surrounds us. This does not mean that it is any less real; the environment, however, isreality constituted in relation to the beings whose environment it is. As I have argued else-where (Ingold 1992a), Gibsonian psychology offers a way of thinking about human-environmental relations that dispenses with the conventional dichotomy between naturallygiven and culturally constructed worlds. According to convention, it is necessary to distin-guish between the ‘real’ environment, as it is presented to detached, scientific observation,and the ‘perceived’ environment as it is built up through a selective response to stimuli(Brookfield 1969: 53). In anthropology, the distinction is commonly expressed by meansof a contrast between the ‘etic’ level of objective description and the ‘emic’ level on whichthe environment is made meaningful by cultural subjects.5 Yet from a Gibsonian perspec-tive, it is apparent that the world becomes a meaningful place for people through beinglived in, rather than through having been constructed along the lines of some formaldesign. Meanings are not attached by the mind to objects in the world, rather these objectstake on their significance – or in Gibson’s terms, they afford what they do – by virtueof their incorporation into a characteristic pattern of day-to-day activities. In short, farfrom being inscribed upon the bedrock of physical reality, meaning is immanent in therelational contexts of people’s practical engagement with their lived-in environments.

PHENOMENOLOGY

It is at this point that ecological psychology makes contact with an older, ContinentalEuropean tradition of philosophical inquiry, broadly characterised as phenomenological,and represented above all in the works of Martin Heidegger and Maurice Merleau-Ponty.Just as the point of departure, for Gibson, had been the perceiver-in-his/her-environment,so likewise these philosophers set out from the premise that every person is, before allelse, a being-in-the-world. And their intellectual agenda, like that of Gibson, was funda-mentally antagonistic to the kind of rationalism whose contemporary manifestation, inthe field of psychology, is cognitive science. Yet in some ways they went even further. Forall his emphasis on perception as a process that is continually going on, Gibson assumedthat the world which the perceiver moves around in and explores is relatively fixed andpermanent, somehow pre-prepared with all its affordances ready and waiting to be takenup by whatever creatures arrive to inhabit it.6 From a phenomenological standpoint, bycontrast, the world emerges with its properties alongside the emergence of the perceiverin person, against the background of involved activity. Since the person is a being-in-the-world, the coming-into-being of the person is part and parcel of the process of coming-into-being of the world.

Consider, for example, Heidegger’s critique of Cartesianism (reviewed in Dreyfus 1991:109–27). Heidegger begins by distinguishing two ways in which the world may ‘show up’to a being who is active within it: availableness and occurrentness. The former is evidentin our everyday use of the most familiar things around us, which, absorbed into the currentof our activity (as indeed, we are ourselves), become in a sense transparent, wholly subor-dinate to the ‘in-order-to’ of the task at hand. The latter refers to the way in which thingsare revealed in their essential nature to an observer who self-consciously stands back fromthe action, assuming a stance of contemplative detachment or disinterested reflection. NowCartesian ontology, which takes as its starting point the self-contained subject confrontinga domain of isolable objects, assumes that things are initially encountered in their pureoccurrentness, or brute facticity. The perceiver has first to make sense of these occurrent

123456789101234567892012345678930123456789401234567118

Dwelling• 168 •

entities – to render them intelligible – by categorising them, and assigning to them mean-ings or functions, before they can be made available for use. Heidegger, however, reversesthis order of priority. For a being whose primary condition of existence is that of dwellingin the world, things are initially encountered in their availableness, as already integratedinto a set of practices for ‘coping’ or getting by. To reveal their occurrent properties,things have to be rendered unintelligible by stripping away the significance they derivefrom contexts of ordinary use. This, of course, is the explicit project of natural science,which seeks to describe and explain a world which the rest of us are preoccupied withliving in. Yet the scientist, like everyone else, is a being-in-the-world, and scientific prac-tice, as any other skilled activity, draws unselfconsciously upon the available. Thus evenscience, however detached and theoretical it may be, takes place against a background ofinvolved activity. The total disengagement of the subject from the world, from whichCartesianism charts a process of building up from the occurrent to the available, is there-fore a pure fiction which can only be reached by extrapolating to the point of absurditya progressive reduction from the available to the occurrent.

If, as Heidegger seems to suggest, self and world merge in the activity of dwelling, sothat one cannot say where one ends and the other begins, it surely follows that the inten-tional presence of the perceiving agent, as a being-in-the-world, must also be an embodiedpresence. This was the principal contention of Merleau-Ponty in his massive treatise, datingfrom 1945 [ trans. 1962], on the Phenomenology of perception. ‘The body’, Merleau-Pontywrote, ‘is the vehicle of being in the world, and having a body is, for a living creature,to be involved in a definite environment, to identify oneself with certain projects and becontinually committed to them’ (1962: 82). Like Heidegger, Merleau-Ponty was concernedto reverse the ontological priorities of Cartesian rationalism. Just as for Heidegger, theavailable is the ground upon which we may seek to reveal the properties of the occurrent,so for Merleau-Ponty our knowledge of the body as a physical thing – as a mere conduitor target of the mind’s attention – is grounded in a more fundamental awareness, pre-objective and pre-conscious, which is given by the existential condition of our total bodilyimmersion, from the start, in an environment. Only because we are thus immersed in theworld can we imagine ourselves as existing separately from it. The problem of perceptionlies in understanding the nature of this immediate pre-objective experience, itself a precon-dition for objective thought. Accordingly, Merleau-Ponty sought to uncover ‘underneaththe objective and detached knowledge of the body that other knowledge which we haveof it by virtue of its always being with us and of the fact that we are our body’ (1962:206, my emphasis). In this latter sense, the body is neither object nor instrument, it israther the subject of perception.

In recent years, albeit somewhat belatedly, many anthropologists have begun to readMerleau-Ponty with renewed interest. Though there is nothing particularly novel aboutanthropological concerns with the body and its symbolism, much work in this field ismarked by a tendency to treat body praxis as a mere vehicle for the outward expression ofmeanings emanating from a higher source in culture or society. This is true, for example,of the writings of Mary Douglas. In line with her general thesis, reviewed in the first part of this chapter, of the cultural construction of experience, Douglas holds that the bodyis a medium whose forms – whether adopted in movement or repose – ‘express social pres-sures in manifold ways’ (1970: 93). As Jackson has eloquently shown, this ‘subjugation ofthe bodily to the semantic’ diminishes the body and its experience in two ways. First, bodymovements – postures and gestures – are reduced to the status of signs which direct theanalyst in search of what they stand for, namely extra-somatic cultural meanings. Secondly,

123456789101231114567892012345678930123456789401234567118


the body is rendered passive and inert, while the active role of mobilising it, putting it touse and charging it with significance is delegated to a knowing subject which is bothdetached from the body and reified as ‘society’ (Jackson 1989: 122–3). The first reductionfails to recognise that gestures, whatever they might be held to symbolise, delineate theirown meanings through their embeddedness in social and material contexts of action. Thesecond reduction ignores a consideration pivotal to Merleau-Ponty’s phenomenology: thatthe body is given in movement, and that bodily movement carries its own immanent intentionality. Indeed it is because of this intentionality that the subject’s action is, at oneand the same time, a movement of perception (1962: 110–11).7

Drawing inspiration from Merleau-Ponty, Jackson (1989) calls for studies that wouldtake as their focus the ‘body subject’ in its dealings with the world. In similar vein, andlinking Merleau-Ponty’s concerns with perception to Bourdieu’s with practice, ThomasCsordas (1990) puts the case for the establishment of a ‘paradigm of embodiment’ inanthropological inquiry. Far from treating the body as an object of study, this paradigmwould be launched from the postulate that ‘the body is to be considered as the subject ofculture, or in other words as the existential [as opposed to the cognitive] ground of culture’(1990: 5). In its promise to collapse the Cartesian dualities between mind and body,subject and object, the paradigm holds a certain appeal for many anthropologists whosefamiliarity with indigenous, non-Western understandings – which are not generally concor-dant with such dualities – predisposes them to adopt a critical attitude towards thefoundational assumptions of Western thought and science. Not everyone has been wonover, however, as is evident from the continuing strength of cognitive anthropology, andfrom the pronouncements of anthropologists such as Bloch (1991), D’Andrade (1995)and Sperber (1996) who see a role for anthropology in an interdisciplinary alliance withcognitive science. Moreover, as I shall show by way of conclusion, there remain threemajor obstacles to the further development of the phenomenological approach.

CONCLUSION

The first obstacle has to do with the problematic status of biology. Even anthropologistswho would readily accept the idea of embodiment as a paradigm for the study of culture,and who denounce the mind/body distinction, tend to balk at attempts to soften theconventional dichotomy between culture and biology (for example, Csordas 1990: 36).In effect, the dichotomy remains as strong as it always was; only the body has been repo-sitioned. Formerly placed with the organism on the side of biology, the body has nowreappeared as a ‘subject’ on the side of culture. Far from collapsing the Cartesian dualismof subject and object, this move actually serves to reproduce it. Moreover it leaves theorganism bodiless, reduced to an inchoate mass of biological potential. The embodimentof culture, in short, leads to nothing less than the disembodiment of the organism! Indeedto posit some kind of biological residuum that exists prior to, and independently of, theculturally constituted body is to resort to the very objectivism that a phenomenologicalapproach claims to repudiate (Morton 1995). It seems to me that to consolidate the theo-retical gains brought by the paradigm of embodiment, one final step has yet to be taken:that is, to recognise that the body is the human organism, and that the process of embod-iment is one and the same as the development of that organism in its environment.

This leads to the second obstacle, which is that the cause of dissolving the divisionbetween body and mind is ill-served by emphasising one term to the exclusion of theother. One could, in principle, speak just as well of enmindment as of embodiment, to

123456789101234567892012345678930123456789401234567118

Dwelling• 170 •

emphasise the way in which the body and its surroundings are incorporated into thoseprocessing loops that underwrite human powers of agency and intentionality. Body andmind, after all, are not two separate things but two ways of describing the same thing –or better, the same process – namely the environmentally situated activity of the humanorganism-person (see Chapter Nineteen, pp. 352–3). Mind, as Gregory Bateson alwaysinsisted, is not ‘in the head’ rather than ‘out there in the world’, but immanent in theactive, perceptual engagement of organism and environment (Bateson 1973). Indeed thedistance between a Merleau-Pontyan phenomenology of the body and what Bateson chris-tened the ‘ecology of mind’ is not as great as might first appear.

Finally, even if it is agreed that a phenomenological approach offers a richer and more‘experience-near’ (Geertz 1984: 124) account of human life in the world than do the moreformal, ‘experience-distant’ concepts of cognitive science, the problem remains of trans-lating this approach into a programme of research that would give us a more accurateidea than we presently have of how people routinely succeed, in their everyday, skilful‘coping’, in performing with ease actions that are far beyond the capabilities of any machineyet devised. It is easy to pour scorn on the efforts of researchers in artificial intelligenceto replicate the processes at work in the human brain, but as Dreyfus admits (1992: xliv),no one knows how the brain does it, nor are philosophers in any way equipped to providethe answers.

What we can say, however, is that the effect of taking the agent-in-an-environmentrather than the isolated, self-contained individual as our point of departure is to collapsenot only the venerable Durkheimian distinction between the individual and society, butalso the division – which has traditionally rested on this distinction – between the twodisciplines of anthropology and psychology. I can see no further intellectual justificationfor continuing to separate these disciplines. For we now recognise that such processes asthinking, perceiving, remembering and learning have to be studied within the ecologicalcontexts of people’s interrelations with their environments. We recognise, too, that themind and its properties are not given in advance of the individual’s entry into the socialworld, but are rather fashioned through a lifelong history of involvement in relationshipswith others. And we know that it is through the activities of the embodied mind (orenminded body) that social relationships are formed and reformed. Psychological and socialprocesses are thus one and the same. And the discipline that will be called into being tostudy these processes, whatever we choose to call it, will be the study of how peopleperceive, act, think, know, learn and remember within the settings of their mutual, prac-tical involvement in the lived-in world.

123456789101231114567892012345678930123456789401234567118


Chapter Ten

Building, dwelling, living:How animals and people make themselves at home in the world

This chapter is partly autobiographical, and describes my own attempts over the last fewyears to find a satisfactory way of understanding the relationships between people andtheir environments. It is incomplete, in the sense that I cannot claim to have yet found,or that I will ever find, final answers to the questions that are bothering me. Indeed, ifone of the main conclusions of what I have to say is that so-called ‘ends’ or ‘goals’ arebut landmarks on a journey, then this must apply as much to my own thinking andwriting as to everything else that people do in the world. The most fundamental thingabout life is that it does not begin here or end there, but is always going on. And for thesame reason, as we saw in Chapter One (p. 20), environments are never complete but arecontinually under construction. My purpose here is to consider the implications of thispoint with regard to our ideas about the similarities and contrasts between human beingsand other animals in the ways in which they go about creating environments for them-selves. I am concerned, in particular, with the meaning of architecture, or of that part ofthe environment which is conventionally described as ‘built’.

In recent years, my own ideas have undergone something of a sea change, which iswhere the autobiographical element comes in. I began with a view that was – and indeedstill is – fairly conventional in anthropology, one that sets out from the premise thathuman beings inhabit discursive worlds of culturally constructed significance, laid outupon the substrate of a continuous and undifferentiated physical terrain. If I differed frommy colleagues, at least in social anthropology, it was in my concern to spell out the impli-cations of this premise for the distinction between human beings and non-human animals.I felt sure that the models developed by ecologists and evolutionary biologists to accountfor the relations between organisms and their environments must apply as well to thehuman as to any other species, yet it was also clear to me that these models left no spacefor what seemed to be the most outstanding characteristic of human activity – that it isintentionally motivated. Human intentions, I argued, are constituted in the intersubjec-tive domain, of relationships among persons, as distinct from the domain in which humanbeings, as biological organisms, relate to other components of the natural environment.Human life, I therefore proposed, is conducted simultaneously in two domains – a socialdomain of interpersonal relations and an ecological domain of inter-organismic relations– so that the problem is to understand the interplay between them (Ingold 1986a: 9).

Starting out from two quite reasonable propositions – that human beings are organ-isms, and that human action is intentionally motivated – I thus ended up with whatappeared to be a thoroughly unreasonable result: that unlike all other animals, humanslive a split-level existence, half in nature, half out; half organism, half person; half body,half mind. I had come out as an unreconstructed Cartesian dualist, which is perhaps not

123456789101234567892012345678930123456789401234567118

Dwelling• 172 •

so surprising when you remember that the intellectual division of labour between thenatural sciences and the humanities – and within anthropology between its biological andsociocultural divisions – rests on a Cartesian foundation. Something, I felt, must be wrongsomewhere, if the only way to understand our own creative involvement in the world isby taking ourselves out of it. Eventually, it dawned on me that although the problem wasan anthropological one, it would require more than an anthropological solution: what isneeded is a completely new way of thinking about organisms and about their relationswith their environments; in short, a new ecology. And it is towards this new ecology thatI have been groping.

In this task, I have gained inspiration from three principal sources. The first comesfrom biology, and consists in the work of the handful of courageous scholars – princi-pally developmental biologists – who have been prepared to challenge the hegemony ofneo-Darwinian thinking in the discipline (e.g. Ho and Saunders 1984, see also Oyama1985). The second lies in what is known as ‘ecological psychology’, an approach to under-standing perception and action that is radically opposed to the cognitivist orientation ofthe psychological mainstream (Gibson 1979, Michaels and Carello 1981). And the thirdcomes from philosophical writing of a broadly phenomenological bent, above all the worksof Martin Heidegger (1971) and Maurice Merleau-Ponty (1962).1 Although developedindependently, in the different disciplinary contexts of biology, psychology and philo-sophy, these three approaches have much in common. Though I cannot now explore thecommonalities in detail, I want to highlight just two of them that are rather central towhat I shall have to say. First, all three approaches reverse the normal order of priority– normal, that is, in the history of Western thought – of form over process. Life, in thisperspective, is not the revelation of pre-existent form but the very process wherein formis generated and held in place. Secondly, the three approaches adopt as their commonpoint of departure the agent-in-its-environment, or what phenomenology calls ‘being inthe world’, as opposed to the self-contained individual confronting a world ‘out there’.In short, they maintain that it is through being inhabited, rather than through its assim-ilation to a formal design specification, that the world becomes a meaningful environmentfor people.

In what follows, I refer to this position as the ‘dwelling perspective’, by contrast to themore conventional position from which I began, and which I shall call the ‘buildingperspective’. Thus the movement in my own thinking has been from the building perspec-tive to the dwelling perspective. To document this movement, I shall start by spelling outthe first of these perspectives, and its implications for the way we understand the construc-tion of the built environment, in greater depth. I shall then explain what is entailed inadopting a dwelling perspective in its place. Finally, I shall consider how this shift froma building perspective to a dwelling perspective bears upon the concept and meaning ofarchitecture.

CONSTRUCTING ENVIRONMENTS AND MAKING WORLDS

Our initial problem may be framed by juxtaposing two statements, the first of which willbe familiar to anthropological readers, the second much less so. ‘Man’, Clifford Geertzhas declared, ‘is an animal suspended in webs of significance he himself has spun’ (1973:5). One is led to suppose that non-human animals are not so suspended. Spiders spinwebs, and do indeed suspend themselves in them, but their webs are tangible objects –they catch flies, not thoughts. But now consider this passage from the delightful but little

123456789101231114567892012345678930123456789401234567118

Building, dwelling, living • 173 •

known text of Jakob von Uexküll, A Strollthrough the Worlds of Animals and Men:‘As the spider spins its threads, everysubject spins his relations to certain char-acters of the things around him, andweaves them into a firm web whichcarries his existence’ (1957: 14). Now thesubjects of which von Uexküll speaks arenot merely human, nor even close tohuman. Indeed he begins his stroll witha particular species of parasitic tick! If, asit would seem, what Geertz says ofhumankind applies equally to ticks, thenwhat – if anything – does distinguishhuman from non-human environments?

Though it might be said, with NelsonGoodman (1978), that human beings aremakers of worlds, this only begs the ques-tion of how human acts of world-makingdiffer from the processes whereby non-human animals fashion their environ-ments. It was this question that initiallyled me to focus on the meaning of thebuilt environment: not, that is, on whata built environment means, but on what it means to say that an environmentis built. How can we distinguish anenvironment that is built from one thatis not? It is all very well to define thebuilt environment, as do Denise Law-rence and Setha Low in a recent review,to include ‘any physical alteration of the natural environment, from hearths

to cities, through construction by humans’ (1990: 454). But why should the products of human building activity be any different, in principle, from the constructionsof other animals? Or to phrase the same question in another way, by what right do weconventionally identify the artificial with the ‘man-made’? And where, in an environ-ment that bears the imprint of human activity, can we draw the line between what is,and is not, a house, or a building, or an instance of architecture (Pearson and Richards1994: 2)?

My first efforts to deal with these questions all hinged on a crucial distinction, whichI thought quite unproblematic at the time, between design and execution. The argumentran roughly as follows: imagine a mollusc shell, a beaver’s lodge and a human house. Allhave been regarded, at one time or another, as instances of architecture. Some authorswould restrict architecture to the house, others would include the lodge – as an exampleof ‘animal architecture’ (von Frisch 1975) – but exclude the shell, others would includeall three forms. The usual argument for excluding the shell is that it is attached to thebody of the mollusc, whereas for something to count as an artefact it must be detached

123456789101234567892012345678930123456789401234567118

Dwelling• 174 •

Figure 10.1 Human and animal architecture. (A) Ground planof beaver lodge (from Morgan 1868: 142); (B) Floor plan andcross-section of Eskimo house, Mackenzie region (from Maussand Beuchat 1979: 4).

from the body. The shell, it is said, ‘just grows’ – there is nothing the mollusc can orneed do about it. The beaver, by contrast, works hard to put its lodge together: the lodgeis a product of the beaver’s ‘beavering’, of its activity. Likewise the house is a product ofthe activities of its human builders. In their respective forms, and levels of complexity,they need not be that different (Figure 10.1). Should we, then, conclude that the lodgeis beaver-made just as much as the house is man-made?

To this question I answered in the negative (Ingold 1986b: 345–6; 1988b: 90).Wherever they are, beavers construct the same kinds of lodges and, so far as we know,have always done so. Human beings, by contrast, build houses of very diverse kinds, andalthough certain house forms have persisted for long periods, there is unequivocal evidencethat these forms have also undergone significant historical change. The difference betweenthe lodge and the house lies, I argued, not in the construction of the thing itself, but inthe origination of the design that governs the construction process. The design of the lodgeis incorporated into the same programme that underwrites the development of the beaver’sown body: thus the beaver is no more the designer of the lodge than is the mollusc thedesigner of its shell. It is merely the executor of a design that has evolved, along with themorphology and behaviour of the beaver, through a process of variation under naturalselection. In other words, both the beaver – in its outward, phenotypic form – and thelodge are ‘expressions’ of the same underlying genotype. Richard Dawkins (1982) hascoined the term ‘extended phenotype’ to refer to genetic effects that are situated beyondthe body of the organism, and in this sense, the lodge is part of the extended phenotypefor the beaver.

Human beings, on the other hand, are the authors of their own designs, constructedthrough a self-conscious decision process – an intentional selection of ideas. As JosephRykwert has put it: ‘unlike even the most elaborate animal construction, human buildinginvolves decision and choice, always and inevitably; it therefore involves a project’ (1991:56). It is to this project, I maintained, that we refer when we say that the house is made,rather than merely constructed. I even went so far as to extend the argument to thedomain of toolmaking, criticising students of animal behaviour for their assumption thatwherever objects are manifestly being modified or constructed for future use, tools arebeing made. They are only being made, I claimed, when they are constructed in the imag-ination prior to their realisation in the material (Ingold 1986a: 40–78). But if the essenceof making lies in the self-conscious authorship of design, that is in the construction of aproject, it follows that things can be made without undergoing any actual physical alter-ation at all. Suppose that you need to knock in a nail but lack a hammer. Looking aroundthe objects in your environment, you deliberately select something best suited to yourpurpose: it must be hard, have a flat striking surface, fit in the hand, and so on. So youpick up an appropriate stone. In this very selection, the stone has ‘become’ a hammer inthat, in your mind’s eye, a ‘hammer-quality’ has been attached to it. Without altering thestone in any way, you have made a hammer out of it.2 In just the same manner, a cavemay come to serve as a dwelling, a stretch of bare flat land as an airstrip, or a shelteredbay as a harbour.

To deal with situations of this kind, I chose the term co-option. Thus the stone wasco-opted, rather than constructed, to become a hammer. It follows that there are twokinds of making: co-optive and constructive. In co-optive making an already existing objectis fitted to a conceptual image of an intended future use, in the mind of a user. Inconstructive making this procedure is reversed, in that the object is physically remodelledto conform more closely to the pre-existing image. Indeed it seemed that the history of

123456789101231114567892012345678930123456789401234567118


things – of artefacts, architecture and landscapes – could be understood in terms of succes-sive, alternating steps of co-option and construction. We press into service what we findaround us to suit our current purposes, we proceed to modify those things to our owndesign so that they better serve these purposes, but at the same time our objectives – oradaptive requirements – also change so that the modified objects are subsequently co-opted to quite other projects for which they are perceived to come in handy, and soon and on. Exactly the same model has been applied to account for the evolution oforganisms – Darwin himself used it in his book on orchids (1862: 348).3 To adopt termssuggested by Stephen J. Gould and Elisabeth Vrba (1982), structures adapted for onepurpose may be exapted for another, subsequently undergoing further adaptation, only tobe exapted for yet another purpose . . . The difference is just that in the case of organicevolution, the selection involved is natural rather than intentional (Ingold 1986b: 200–2).

It was in searching around for ways to express these ideas that I came across the writ-ings of Jakob von Uexküll, Estonian-born aristocrat and a founding figure in the fieldsof both ethology and semiotics, to whose Stroll through the Worlds of Animals and Men,first published in 1934, I have already referred. Reacting against the mechanistic biologyof the day, von Uexküll argued that to treat the animal as a mere assemblage of sensoryand motor organs is to leave out the subject who uses these organs as tools, respectively,of perception and action:

We who still hold that our sense organs serve our perceptions, and our motor organsour actions, see in animals . . . not only the mechanical structure, but also the oper-ator, who is built into their organs, as we are into our bodies. We can no longer regardanimals as mere machines, but as subjects whose essential activity consists in perceivingand acting . . . All that a subject perceives becomes his perceptual world and all that hedoes, his effector world. Perceptual and effector worlds together form a closed unit, theUmwelt.

(1957: 6)

For von Uexküll, the Umwelt – that is, the world as constituted within the specific lifeactivity of an animal – was to be clearly distinguished from the environment, by whichhe meant the surroundings of the animal as these appear to the indifferent human observer.We human beings cannot enter directly into the Umwelten of other creatures, but throughclose study we may be able to imagine what they are like. But the reverse does not hold:the non-human animal, because it cannot detach its consciousness from its own life-activity, because it is always submerged within its own Umwelt, cannot see objects as such,for what they are in themselves. Thus for the animal, the environment – conceived as adomain of ‘neutral objects’ – cannot exist (Ingold 1992a: 43).

Towards the end of his stroll, von Uexküll invites his readers to imagine the manifoldinhabitants of an oak tree. There is the fox, who has built its lair between the roots; theowl, who perches in the crotch of its mighty limbs; the squirrel, for whom it provides averitable maze of ladders and springboards; the ant, who forages in the furrows and cragsof its bark; the wood-boring beetle who feeds and lays its eggs in passages beneath thebark, and hundreds of others (Figures 10.2 and 10.3). Each creature, through the sheerfact of its presence, confers on the tree – or on some portion of it – a particular qualityor ‘functional tone’: shelter and protection for the fox, support for the owl, a thorough-fare for the squirrel, hunting grounds for the ant, egg-laying facilities for the beetle. Thesame tree, thus, figures quite differently within the respective Umwelten of its diverse

123456789101234567892012345678930123456789401234567118

Dwelling• 176 •

inhabitants. But for none of them does itexist as a tree (von Uexküll 1957: 76–9).Now consider the forester, who is measuringup the tree to estimate the volume of timberit will yield. For him, the tree figures as apotential source of valuable raw material,whereas for the little child – again to followvon Uexküll’s example (pp. 73–5) – it seemsto be alive and to reveal a frightening aspect.But these different perceptions are not tied,as they are for non-human animals, to themodus operandi of the organism. Humanbeings do not construct the world in acertain way by virtue of what they are, butby virtue of their own conceptions of thepossibilities of being. And these possibilitiesare limited only by the power of the imag-ination.

Herein, it seemed to me, lay the essen-tial distinction I was seeking between therespective ways in which the subjectiveexistence of human and non-human animalsis suspended in ‘webs of significance’. Forthe non-human, every thread in the web isa relation between it and some object orfeature of the environment, a relation thatis set up through its own practical im-mersion in the world and the bodilyorientations that this entails. For thehuman, by contrast, the web – and the rela-tions of which it consists – are inscribed ina separate plane of mental representations,forming a tapestry of meaning that coversover the world of environmental objects. Whereas the non-human animal perceives theseobjects as immediately available for use, to human beings they appear initially as occur-rent phenomena to which potential uses must be affixed, prior to any attempt atengagement. The fox discovers shelter in the roots of a tree, but the forester sees timberonly in his mind’s eye, and has first to fit that image in thought to his perception of theoccurrent object – the tree – before taking action. Or to take another example, suggestedrecently by Maurice Bloch, the ‘swidden plot’ exists as an image in the mind of the horti-culturalist, who has to match that image to an observed stand of uncut forest prior totransforming it into a field (Bloch 1991: 187). As mental representations, the timber andthe swidden plot belong to the ‘intentional worlds’ (cf. Shweder 1990: 2) of the foresterand the farmer; as occurrent phenomena, the oak tree and the stand of forest belong tothe physical environment of ‘neutral objects’. It has been conventional, in anthropo-logical and other writings of Western academic provenance, to refer to these worlds, ofhuman values and purposes on the one hand, and of physical objects on the other, bymeans of the shorthand terms, culture and nature, respectively. And in a paper written

123456789101231114567892012345678930123456789401234567118


Figure 10.2 Fox, owl and oak tree

From Jakob von Uexküll ‘A Stroll through the Worlds ofAnimals and Men,’ in Instinctive Behavior, 1957, pp. 76–7,illustrations by G. Kriszat.

in 1987, I concluded that ‘making is equivalent to the cultural ordering of nature – theinscription of ideal design upon the material world of things’ (Ingold 1989: 506). Thisstatement, I confess, is now a source of considerable embarrassment.

THE BUILDING PERSPECTIVE

In my defence, I can only say that I was singing a tune that has been sung by mostanthropologists, in one form or another, for decades, in the context of an encounter withstudents of animal behaviour whose theories had no place for agency or intentionality atall, except as an epiphenomenal effect of innate predisposition.4 This tune is what I earliercalled the ‘building perspective’, and I should now like to elaborate on this perspectivewith reference to anthropological work other than my own. For a founding statement, wecould turn once again to Geertz, and to his assertion that culture – or at least that kindof culture taken to be the hallmark of humanity – consists in ‘the imposition of an arbi-trary framework of symbolic meaning upon reality’ (1964: 39). Reality, that which isimposed upon, is envisioned here as an external world of nature, a source of raw mate-rials and sensations for diverse projects of cultural construction. Following from this, adistinction is commonly made between the real environment that is given independentlyof the senses, and the perceived environment as it is reconstructed in the mind throughthe ordering of sense data in terms of acquired, cognitive schemata. Other conventionaloppositions that encode the same distinction, and that we have already encountered (seeChapter Three, p. 41, and Chapter Nine, p. 168), are between ‘etic’ and ‘emic’, andbetween ‘operational’ and ‘cognised’. The starting point in all such accounts is an imaginedseparation between the perceiver and the world, such that the perceiver has to reconstructthe world, in the mind, prior to any meaningful engagement with it.

123456789101234567

8920123

45678930123456789401234567118

Dwelling• 178 •

Figure 10.3 Ant, bark-boring beetle and oak tree

From Jakob von Uexküll ‘A Stroll through the Worlds of Animals and Men,’ in Instinctive Behavior,1957, pp. 78–9, illustrations by G. Kriszat.

Here, then, is the essence of the building perspective: that worlds are made before theyare lived in; or in other words, that acts of dwelling are preceded by acts of worldmaking.A good example of this approach comes from the introduction to Maurice Godelier’sbook, The mental and the material (1986). Here, Godelier is concerned with the propertranslation of the Marxian concepts Grundlage and Überbau, usually rendered in Englishas ‘infrastructure’ and ‘superstructure’. He likens the Überbau to a building: ‘The Überbauis a construction, an edifice which rises on foundations, Grundlage; and it [the Überbau]is the house we live in, not the foundations’ (pp. 6–7). Human beings, then, inhabit thevarious houses of culture, pre-erected upon the universal ground of nature – includingthe universals of human nature. For another example, I would like to turn to Peter Wilson’sThe domestication of the human species (1988). In this book, Wilson argues that the mostsignificant turning point in human social evolution came at the moment when peoplebegan to live in houses. Roughly speaking, this marks a division between hunters andgatherers, on the one hand, and agriculturalists and urban dwellers, on the other. ‘Hunter-gatherers’, Wilson writes, ‘create for themselves only the flimsiest architectural context,and only the faintest line divides their living space from nature’. All other societies, bycontrast, ‘live in an architecturally modified environment’, inhabiting houses and villagesof a relatively enduring kind, structures that – even when abandoned – leave an almostindelible impression in the landscape. In essence, Wilson is distinguishing between soci-eties with architecture and societies without it.

This is a bold generalisation, and like all such, it is an easy target for empirical refutation.That is not my concern, however. I am rather concerned to expose the assumptions entailedin making the distinction between an ‘architecturally modified environment’ and what is

123456789101231114567892012345678930123456789401234567118


Figure 10.4 The Mbuti Pygmy camp of Apa Lelo

From C. M. Turnbull, Wayward servants, published by Eyre & Spottiswoode, 1965, p. 357.

simply called ‘nature’. For it is on thisdistinction that Wilson’s entire argumentrests. One objection to it immediatelycomes to mind. To be sure, the physicalarrangement and formal properties of ahunter-gatherer encampment may be verydifferent from those of a permanent villagesettlement. By way of example, compare theplan, shown in Figure 10.4, of the MbutiPygmy camp of Apa Lelo, in the Ituri forestof Zaire, with the plans shown in Figure10.5 of the ancient Mesopotamian villagesite of Tell es-Sawwan. In the first case thespatial structure of settlement is loose,informal, and sensitive to the changing stateof interpersonal relations between cliques,hosts and visitors. In the second it is tightlypacked, geometrically regular, and appearsto impose fairly tight constraints on the dis-position of people and activities. Moreover,compared with the substantial buildings ofthe village settlement, the constructions ofthe hunter-gatherers are scarcely more thatshades and windbreaks. Most of life, forhunter-gatherers, goes on around dwellingsrather than in them. Nevertheless, the factremains that hunter-gatherers do build shel-ters of various kinds. So who are we to saythat they have no architecture? And if theydo not, how are we to comprehend theirbuilding activity?

The answer that emerges from Wilson’saccount is that among hunter-gatherers,erecting shelters is one of a suite of activi-

ties, along with food-collecting, cooking, toolmaking and repair, childminding, and so on,that constitute the daily round for these people. Thus building activity is part and parcelof life in an environment that is already given in nature, and that has not itself been arti-ficially engineered. With village architecture, by contrast, nature has to a degree beencovered over or transformed, so that what immediately confronts people is not a naturalenvironment but – in Wilson’s words – ‘an environment of their own making, the cultural’(1988: 8). If hunter-gatherers build as part of their adaptation to the given conditions ofthe natural environment, villagers adapt to the conditions of an environment that is alreadybuilt. Either way, the environment is given in advance, as a kind of container for life tooccupy. Where, as among hunter-gatherers, building is a part of everyday life, it is notsupposed to have any lasting impact on the environment; where, as among villagers, theenvironment has been manifestly built, the buildings are apparently made before life beginsin them. This, of course, is the architect’s perspective: first plan and build the houses,then import the people to occupy them.

123456789101234567892012345678930123456789401234567118

Dwelling• 180 •

Figure 10.5 Building plans of three periods from the ancientMesopotamian site of Tell es-Sawwan.

From J. Mellaart, The Neolithic of the Near East, published byThames and Hudson, London 1975.

What, then, of the dwellings of nomadic pastoralists? A recent study comparing pastoraltent dwellings and village houses in Turkey and Iran by the archaeologist, Roger Cribb(1991), found that despite differences in the building materials used and the flexibilitythey afford, the tent and the house were virtually identical in their underlying organisa-tional templates. What really distinguished the house from the tent was the degree towhich the imposed, cultural design – shared by villagers and nomads alike – is actuallytranslated into enduring, material structures. For such structures do not get built overnight;they grow cumulatively in the course of a settlement’s continuous occupation, such that‘each new alteration or addition builds on a series of existing structures’. But in the caseof a pastoral nomadic camp, ‘each occupation is a fresh event’, so that the camp ‘has nosuch history but remains permanently retarded in the initial stages of the normal devel-opmental cycle [of the settlement]’ (1991: 156). Thus, although pastoralists carry a basicorganisational template with them, there is little opportunity for its enduring physical real-isation before the camp picks up and moves off somewhere else, where the occupationprocess starts all over again. In such cases, building never proceeds beyond the first phaseof temporary habitation (Ingold 1992c: 795–6).

In a statement that epitomises the building perspective, Amos Rapoport writes that ‘theorganisation of space cognitively precedes its material expression; settings and built envi-ronments are thought before they are built’ (1994: 488). In the case of villagers, theenvironment is ready-built. In the case of nomadic pastoralists, it would seem, the environ-ment, though thought, is never more than partially built. As for the hunter-gatherers, it appears that the building hardly gets started at all: indeed Rapoport refers to the camp sites of Aboriginal people of the Australian Central Desert as exemplars of the situ-ation where the environment is thought but never built. On these grounds, as we saw inChapter Three (pp. 56–7), they are supposed to inhabit a ‘natural’ rather than an ‘artificial’environment.

THE SEARCH FOR ORIGINS

Having spelled out the essence of the building perspective, let me now return to my earlierobservation, comparing the forms of the beaver’s lodge and the human house, that thefirst is tied, as it were, to the nature of the beaver itself, whereas the second is both histor-ically and regionally variable. Among non-human animals, it is widely supposed, therecan be no significant change in built form that is not bound to evolutionary changes inthe essential form of the species. With human beings, by contrast, built form is free tovary independently of biological constraint, and to follow developmental pathways of itsown, effectively decoupled from the process of evolution. In his famous paper of 1917,on ‘The Superorganic’, Alfred Kroeber declared: ‘Who would be so rash as to affirm thatten thousand generations of example would convert the beaver from what he is into acarpenter or a bricklayer – or, allowing for his physical deficiency in the lack of hands,into a planning engineer!’ (1952: 31). Yet human beings, through practice, example anda good measure of ingenuity, coupled with their ability to transmit their acquired know-how across the generations and to preserve it in long-term memory, have learned all thesetrades, and many more besides.

However, this argument implies some kind of threshold in the evolution of our ownkind, at which point our ancestors were sufficiently endowed with the qualities of intel-ligence and manual dexterity to become the authors of their own projects of building.Taking off from this point, the history of architecture must be supposed to have proceeded

123456789101231114567892012345678930123456789401234567118


from the earliest dwellings to the modern construction industry, the species-specific natureof the human organism remaining all the while unchanged. But what was the earliestdwelling? According to Kenneth Bock, an event in the history of architecture – such asthe construction of a Gothic vault – differs from an event in the evolution of species ‘inthat the former involves formation of intent or purpose on the part of an actor while thelatter does not’ (1980: 182). The same idea is implied by Joseph Rykwert when he suggeststhat the essence of architecture lies in ‘taking thought about building’ (1991: 54). Buthow did it come about that, at some decisive moment, our ancestors began to think aboutwhat they built?

As Rykwert shows, in his study of the notion of the ‘primitive hut’ in the history ofarchitecture, this is a question that has long exercised the minds of Western thinkers. Andthe title of his book, On Adam’s House in Paradise (1972), nicely conveys the mythicquality of the many speculative answers that have been proposed. Reproduced in Figure10.6 is one of the more delightful images of ‘the first hut’, taken from the work of thegreat French architectural theorist, Eugène Viollet-le-Duc, Histoire de l’habitation humaine,published in 1875 (Viollet-le-Duc 1990: 26). Architecture began, for Viollet-le-Duc, whenthe problem of the need for shelter was met through the procedures of rational planning.In his tale of the building of the first hut the secret is revealed to a hapless primitive tribe,the Nairitti, by a progressive time-traveller by the name of Epergos, bestowed upon themas a gift of his superior intelligence. For Viollet-le-Duc, as for many others, Rykwert notes,it was ‘the difference of conception, the attachment of meaning to his task, that distin-guishes man’s first attempts [at building] from those of the instinctually driven beasts’(1972: 22). These attempts may have been decidedly inferior to the constructions ofanimals, nevertheless they marked the turning point at which humanity was set upon theroad to culture and civilisation.

The search for the first building continues to this day, though it is informed by a muchbetter knowledge both of the archaeological traces left by early human or hominid popu-lations, and of the behaviour of those species of animals – namely the great apes – mostclosely related to humankind. One of the most peculiar and distinctive aspects of thebehaviour of chimpanzees, gorillas and orang-utans is their habit of building so-called‘nests’. In functional terms, they are not really nests at all: every individual animal buildsits own nest afresh, each evening, and uses it for the sole purpose of sleeping. Nor doesthe nest site mark any kind of fixed point in the animal’s movements; it may be builtanywhere, and is abandoned the next morning (Groves and Sabater Pi 1985: 23).Nevertheless, assuming that the common ancestor of apes and humans would have hada similar habit, attempts have been made to trace an evolutionary continuum from thisnesting behaviour to the residential arrangements of prototypical human groups (of whichthe camps of contemporary hunter-gatherers have frequently been taken as the closestexemplars, on the grounds of the presumed similarity of ecological context).

Comparing the nesting patterns of apes with the camping patterns of human hunter-gatherers, Colin Groves and J. Sabater Pi note some striking differences. The human ‘nest’,if we may call it that, is a fixed point for the movements of its several occupants, and aplace to which they regularly return. In other words, it has the attributes of what theethologist, Heini Hediger, would call ‘home’: it is a ‘goal of flight’ and a ‘place of maximalsecurity’ (Hediger 1977: 181). There is a difference, too, in the respective ways in whichapes and humans go about building their accommodation. For one thing, apes use materialthat comes immediately to hand, normally by a skilful interweaving of growing vegeta-tion to form an oval-shaped, concave bed; whereas humans collect suitable materials from

123456789101234567892012345678930123456789401234567118

Dwelling• 182 •

a distance, prior to their assembly into a convex, self-supporting structure. For anotherthing, the ape makes its nest by bending the vegetation around its own body; whereasthe human builds a hut, and then enters it (Groves and Sabater Pi 1985: 45). There isa sense, as Hediger remarks, in which apes build from the ‘bottom up’, seeking supportfor rest and sleeping, whereas humans build from the ‘top down’ seeking shelter fromsun, rain or wind (1977: 184). Yet there are also remarkable similarities between ape andhuman living arrangements, in the overall number and layout of nests or huts and in theunderlying social organisation, and on the grounds of these similarities, Groves and SabaterPi feel justified in arguing that human campsites are but elaborations of a generalised apepattern. All the critical differences – the functioning of the site as a home-base, the collec-tion of material prior to construction, the technique of building from the outside – canbe put down, they think, to one factor, namely the human ability ‘to visualise objects innew configurations, and to bring these configurations into being on the basis of thatmental picture’ (1985: 45).

123456789101231114567892012345678930123456789401234567118


Figure 10.6 The first hut, as depicted by Viollet-le-Duc.

From The architectural theory of Viollet-le-Duc: readings and commentary, edited by M. F. Hearn, publishedby MIT Press, 1990, p. 26.

Though in substance based on fact rather than fantasy, the form in which this argu-ment is cast is virtually identical to that of Viollet-le-Duc’s tale of the building of thefirst hut. Equipped, albeit by natural selection rather than providential intervention, withforesight and intelligence, the first builders set to work to execute a plan that was alreadyformed as a picture in their imagination. They had solved the problem of shelter in theirminds, prior to putting the solution into practical effect. It is in this light that we canunderstand the extraordinary significance that has been attached to the so-called ‘stonecircle’ discovered at the famous site of Olduvai Gorge in Tanzania, and dated to some1.75 million years ago (Figure 10.7). In her interpretation of the circle, Mary Leakeywrites that in its general appearance, it ‘resembles temporary structures often made bypresent-day nomadic peoples who build a low stone wall round their dwellings to serveeither as a windbreak or as a base to support upright branches which are bent over andcovered with either skins or grass’ (1971: 24). A photograph of such a dwelling, from theOkombambi people of Southwest Africa, is provided to substantiate the comparison. Asalways in these matters, the specific interpretation has been challenged. What has not beenchallenged, however, is the frame of mind that leads us to suppose that if the interpre-tation were correct, we would have at last discovered the real ‘first hut’, and with it notjust the origins of architecture, but the point of transition to true humanity.

1234567891012345678920123456789

30123456789401234567118

Dwelling• 184 •

Figure 10.7 The ‘stone circle’ from Bed I of Olduvai Gorge.

From M. D. Leakey, Olduvai Gorge (volume three), published by Cambridge University Press, 1971.

For it is the structure of our thought, not the patterning of the archaeological record, thatsets up a point of origin at the intersection of two axes, one of evolutionary change – leadingfrom ancestral pongids and hominids to human beings, the other of historical change – lead-ing from Palaeolithic hunting and gathering to modern industry. (Why this should be so is amatter to which I return in Chapter Twenty-one, pp. 388–90.) To explode the myth of thefirst hut thus requires nothing less than the dissolution of the dichotomy, which in modernscholarship separates the biological sciences from the humanities, between evolution andhistory, or between the temporal processes of nature and culture. Before indicating how thismight be done, I need to introduce what I have called the ‘dwelling perspective’.

THE DWELLING PERSPECTIVE

For this purpose I turn to Martin Heidegger’s evocative essay, ‘Building DwellingThinking’, on which I have drawn for my title (Heidegger 1971: 145–61). In this essay,Heidegger asks what it means to build and to dwell, and what the relation is betweenthese two – between building and dwelling. He begins with what might be taken as thehegemonic view, as enshrined in the discourse of Western modernity. This is that buildingand dwelling are separable but complementary activities, related as means to ends. Webuild houses so that we may dwell in them (or, as is usual in industrial society, somepeople build houses for other people to live in). To dwell, in this sense, means merely‘to occupy a house, a dwelling place’. The building is a container for life activities, ormore strictly for certain life activities, since there are other kinds of activity that go onoutside houses, or in the open air. Yet, Heidegger asks, ‘do the houses in themselves holdany guarantee that dwelling occurs in them?’ (1971: 146). To clarify matters, let us callthe physical structure, the building in itself, the house; and the setting within which peopledwell the home (Lawrence 1987). Heidegger’s question can then be rephrased as follows:what does it take for a house to be a home (Pearson and Richards 1994: 6)? Merely topose the question in this form suggests that there must be more to dwelling than themere fact of occupation. What, then, does it mean, ‘to dwell’?

Heidegger tackles the issue through an exercise in etymology. The current German wordfor the verb ‘to build’, bauen, comes from the Old English and High German buan,meaning ‘to dwell’. Though this original meaning has been lost, it is preserved in suchcompounds as the English ‘neighbour’, meaning one who dwells nearby. Moreover, thissense of dwelling was not limited to one sphere of activity among many – to domesticlife, say, as opposed to work or travel. Rather it encompassed the whole manner in whichone lives one’s life on the earth; thus ‘I dwell, you dwell’ is identical to ‘I am, you are’.Yet bauen has another sense: to preserve, to care for, or more specifically to cultivate orto till the soil. And then there is the third sense: to construct, to make something, toraise up an edifice. Both these modern senses of building – as cultivation and as construc-tion – are thus shown to be encompassed within the more fundamental sense of dwelling.In the course of time, however, this underlying sense has fallen into disuse, such thatbauen has come to be reserved exclusively for cultivation and construction. Havingforgotten how the latter activities are grounded in dwelling, modern thought then redis-covers dwelling as the occupation of a world already built.

In short, where before, building was circumscribed within dwelling, the position nowappears reversed, with dwelling circumscribed within building. Heidegger’s concern is toregain that original perspective, so that we can once again understand how the activities ofbuilding – of cultivation and construction – belong to our dwelling in the world, to the

123456789101231114567892012345678930123456789401234567118


way we are. ‘We do not dwell because we have built, but we build and have built becausewe dwell, that is because we are dwellers . . . To build is in itself already to dwell . . . Onlyif we are capable of dwelling, only then can we build ’ (Heidegger 1971: 148, 146, 160, orig-inal emphases). I take this to be the founding statement of the dwelling perspective.5 Whatit means is that the forms people build, whether in the imagination or on the ground, arisewithin the current of their involved activity, in the specific relational contexts of theirpractical engagement with their surroundings. Building, then, cannot be understood as asimple process of transcription, of a pre-existing design of the final product onto a rawmaterial substrate. It is true that human beings – perhaps uniquely among animals – havethe capacity to envision forms in advance of their implementation, but this envisioning isitself an activity carried on by real people in a real-world environment, rather than by adisembodied intellect moving in a subjective space in which are represented the problemsit seeks to solve (see Chapter Twenty-three, pp. 418–19). In short, people do not importtheir ideas, plans or mental representations into the world, since that very world, to borrowa phrase from Merleau-Ponty (1962: 24), is the homeland of their thoughts. Only becausethey already dwell therein can they think the thoughts they do.

To argue that the forms of buildings arise as a kind of crystallisation of human activitywithin an environment clearly puts paid to my initial dichotomy between design andexecution. No longer can we assume, with Christopher Alexander, that form is ‘the ultim-ate object of design’ (1964: 15), as though the one issued quite automatically andunproblematically from the other. To the contrary, a dwelling perspective ascribes thegeneration of form to those very processes whose creativity is denied by that perspectivewhich sees in every form the concrete realisation of an intellectual solution to a designproblem. Where, then, does this leave the constructions of non-human animals? The argu-ment is equally damaging to the conventional biological account, which holds that theoutward, phenotypic form – not just of the animal itself, but of the constructions makingup its ‘extended phenotype’ – is the expression of a solution to some specific problem ofadaptation, already reached by natural selection, and transferred to the animal at the pointof conception, encrypted in the materials of heredity – the genes. That design is thusimported into the organism, as a kind of ‘evolved architecture’ (Tooby and Cosmides1992), prior to the organism’s development within an environmental context, is indeedone of the great delusions of modern biology. For as I shall show in Chapter Twenty-one, the forms of organisms are in no way prefigured in their genes but are the emergentoutcomes of environmentally situated development processes.

For any animal, the environmental conditions of development are liable to be shapedby the activities of predecessors. The beaver, for example, inhabits an environment thathas been decisively modified by the labours of its forbears, in building dams and lodges,and will in turn contribute to the fashioning of an environment for its progeny. It is insuch a modified environment that the beaver’s own bodily orientations and patterns ofactivity undergo development. The same goes for human beings. Human children, likethe young of many other species, grow up in environments furnished by the work ofprevious generations, and as they do so they come literally to carry the forms of theirdwelling in their bodies – in specific skills, sensibilities and dispositions. But they do notcarry them in their genes, nor is it necessary to invoke some other kind of vehicle for theinter-generational transmission of information – cultural rather than genetic – to accountfor the diversity of human living arrangements.

We can now see how, by adopting a dwelling perspective – that is, by taking the animal-in-its-environment rather than the self-contained individual as our point of

123456789101234567892012345678930123456789401234567118

Dwelling• 186 •

departure – it is possible to dissolve the orthodox dichotomies between evolution andhistory, and between biology and culture. For if, by evolution, we mean differentiationover time in the forms and capacities of organisms, then we would have to admit that changes in the bodily orientations and skills of human beings, insofar as they arehistorically conditioned by the work of predecessors (along with the enduring products of that work, such as buildings), must themselves be evolutionary. And if, by culturalvariation, we mean those differences of embodied knowledge that stem from the diversityof local developmental contexts, then far from being superimposed upon a substrate of evolved human universals, such variation must be part and parcel of the variation ofall living things, which has its source in their enmeshment within an all-encompassingfield of relations. It is not necessary, then, to invoke one kind of theory, of biologicalevolution, to account for the transition from nest to hut, and another kind, of culturalhistory, to account for the transition from hut to skyscraper. For once history is itselfrecognised as an evolutionary process, the point of origin constituted by the intersectionof evolutionary and historical continua disappears, and the search for the first hut – for the beginnings of architecture, history and true humanity – becomes a quest after anillusion.6

THE HOUSE AS ORGANISM

Let me conclude by returning to von Uexküll’s oak tree. Suppose that it stands, not inthe forest, but in the precincts of a house. Now at first glance we might have no hesita-tion in regarding the house, but not the tree, as a building, or an instance of architecture.For surely the house, as Godelier puts it, belongs to ‘that part of nature which is trans-formed by human action and thought [and] owes its existence to conscious human actionon nature’ (1986: 5, see also Chapter Five p. 79). The tree, on the other hand, has nosuch debt to humanity, for it has grown there, rooted to the spot, entirely of its ownaccord. On closer inspection, however, this distinction between those parts of the environ-ment that are, respectively, built and unbuilt seems far less clear. For the form of the treeis no more given, as an immutable fact of nature, than is the form of the house an impo-sition of the human mind. Recall the many inhabitants of the tree: the fox, the owl, thesquirrel, the ant, the beetle, among countless others. All, through their various activitiesof dwelling, have played their part in creating the conditions under which the tree, overthe centuries, has grown to assume its particular form and proportions. And so, too, havehuman beings, in tending the tree’s surroundings.

But the house also has many and diverse animal inhabitants – more, perhaps, than weare inclined to recognise. Sometimes special provision is made for them, such as the kennel,stable or dovecote. Others find shelter and sustenance in its nooks and crannies, or evenbuild there. And all, in their various ways, contribute to its evolving form, as do thehouse’s human inhabitants in keeping it under repair, decorating it, or making structuralalterations in response to their changing domestic circumstances. Thus the distinctionbetween the house and the tree is not an absolute but a relative one – relative, that is,to the scope of human involvement in the form-generating process.7 Houses, as SuzanneBlier notes (1987: 2), are living organisms. Like trees, they have life-histories, which consistin the unfolding of their relations with both human and non-human components of theirenvironments. To the extent that the influence of the human component prevails, anyfeature of the environment will seem more like a building; to the extent that the non-human component prevails, it will seem less so.

123456789101231114567892012345678930123456789401234567118


Building, then, is a process that is continually going on, for as long as people dwell inan environment. It does not begin here, with a pre-formed plan, and end there, with afinished artefact. The ‘final form’ is but a fleeting moment in the life of any feature, whenit is matched to a human purpose, likewise cut out from the flow of intentional activity.As the philosopher Alfred North Whitehead once remarked, ‘from the moment of birthwe are immersed in action, and can only fitfully guide it by taking thought’ (1938: 217).And this applies, with equal force, to ‘taking thought about building’, the definitive char-acteristic of the architectural attitude. We may indeed describe the forms in ourenvironment as instances of architecture, but for the most part we are not architects. Forit is in the very process of dwelling that we build.

123456789101234567892012345678930123456789401234567118

Dwelling• 188 •

Chapter Eleven

The temporality of the landscape

PROLOGUE

I adhere to the view that social or cultural anthropology, biological anthropology andarchaeology form a necessary unity – that they are all part of the same intellectual enter-prise. I am not concerned here with the link with biological or ‘physical’ anthropology,but what I have to say does bear centrally on the unifying themes of archaeology andsociocultural anthropology. I want to stress two such themes, and they are closely related.First, human life is a process that involves the passage of time. Secondly, this life-processis also the process of formation of the landscapes in which people have lived. Time andlandscape, then, are to my mind the essential points of topical contact between archae-ology and anthropology. My purpose, in this chapter, is to bring the perspectives ofarchaeology and anthropology into unison through a focus on the temporality of thelandscape. In particular, I believe that such a focus might enable us to move beyond thesterile opposition between the naturalistic view of the landscape as a neutral, external back-drop to human activities, and the culturalistic view that every landscape is a particularcognitive or symbolic ordering of space. I argue that we should adopt, in place of boththese views, what I have called a ‘dwelling perspective’, according to which the landscapeis constituted as an enduring record of – and testimony to – the lives and works of past generations who have dwelt within it, and in so doing, have left there something ofthemselves.

For anthropologists, to adopt a perspective of this kind means bringing to bear theknowledge born of immediate experience, by privileging the understandings that peoplederive from their lived, everyday involvement in the world. Yet it will surely be objectedthat this avenue is not open to archaeologists concerned with human activities in thedistant past. ‘The people’, it is said, ‘they’re dead’ (Sahlins 1972: 81); only the materialrecord remains for their successors of our own time to interpret as best they can. But thisobjection misses the point, which is that the practice of archaeology is itself a form of dwelling.The knowledge born of this practice is thus on a par with that which comes from thepractical activity of the native dweller and which the anthropologist, through participa-tion, seeks to learn and understand. For both the archaeologist and the native dweller,the landscape tells – or rather is – a story, ‘a chronicle of life and dwelling’ (Adam 1998:54). It enfolds the lives and times of predecessors who, over the generations, have movedaround in it and played their part in its formation. To perceive the landscape is there-fore to carry out an act of remembrance, and remembering is not so much a matter of calling up an internal image, stored in the mind, as of engaging perceptually with anenvironment that is itself pregnant with the past. To be sure, the rules and methods of

123456789101231114567892012345678930123456789401234567118


engagement employed respectively by the native dweller and the archaeologist differ, asdo the stories they tell. Nevertheless, insofar as both seek the past in the landscape, theyare engaged in projects of fundamentally the same kind.1

It is of course part of an archaeological training to learn to attend to those clues whichthe rest of us might pass over (literally, when they are below the surface), and which makeit possible to tell a fuller or a richer story. Likewise native dwellers, along with theiranthropological companions, learn through an education of attention. The novice hunter,for example, travels through the country with his mentors, and as he goes, specific featuresare pointed out to him. Other things he discovers for himself, in the course of furtherforays, by watching, listening and feeling. Thus the experienced hunter is the knowledge-able hunter (see Chapter Three, pp. 55–6). He can tell things from subtle indications thatyou or I, unskilled in the hunter’s art, might not even notice. Called upon to explicatehis knowledge, he may do so in a form that reappears in the work of the non-nativeethnographer as a corpus of myths or stories, whereas the archaeologist’s knowledge –drawn from the practices of excavation rather than hunting – may appear in the seem-ingly authoritative form of the site report. But we should resist the temptation to assumethat since stories are stories they are, in some sense, unreal or untrue, for this is to supposethat the only real reality, or true truth, is one in which we, as living, experiencing beings,can have no part at all. Telling a story, as I observed in Chapter Three (p. 56), is notlike unfurling a tapestry to cover up the world, it is rather a way of guiding the attentionof listeners or readers into it. A person who can ‘tell’ is one who is perceptually attunedto picking up information in the environment that others, less skilled in the tasks ofperception, might miss, and the teller, in rendering his knowledge explicit, conducts theattention of his audience along the same paths as his own.

Following that preamble, I shall now go on to lay out the burden of my argument.This is presented in four principal sections. In the first two, I attempt to specify moreprecisely what I mean by my key terms – landscape and temporality. I argue that tempo-rality inheres in the pattern of dwelling activities that I call the taskscape. In the thirdsection I consider how taskscape relates to landscape and, ultimately by dissolving thedistinction between them, I proceed to recover the temporality of the landscape itself.Finally, I draw some concrete illustrations of my arguments from a well-known paintingby Bruegel, The harvesters.

LANDSCAPE

Let me be begin by explaining what the landscape is not. It is not ‘land’, it is not ‘nature’,and it is not ‘space’. Consider, first of all, the distinction between land and landscape. Landis not something you can see, any more than you can see the weight of physical objects. All objects of the most diverse kinds have weight, and it is possible to express how muchanything weighs relative to any other thing. Likewise, land is a kind of lowest commondenominator of the phenomenal world, inherent in every portion of the earth’s surface yetdirectly visible in none, and in terms of which any portion may be rendered quantitativelyequivalent to any other (Ingold 1986a: 153–4).2 You can ask of land, as of weight, howmuch there is, but not what it is like. But where land is thus quantitative and homogeneous,the landscape is qualitative and heterogeneous. Supposing that you are standing outdoors,it is what you see all around: a contoured and textured surface replete with diverse objects– living and non-living, natural and artificial (these distinctions are both problematic, as weshall see, but they will serve for the time being). Thus at any particular moment, you can

123456789101234567892012345678930123456789401234567118

Dwelling• 190 •

ask of a landscape what it is like, but not how much of it there is. For the landscape is aplenum, there are no holes in it that remain to be filled in, so that every infill is in realitya reworking. As Meinig observes, one should not overlook ‘the powerful fact that life mustbe lived amidst that which was made before’ (1979a: 44).

The landscape is not ‘nature’. Of course, nature can mean many things, and this is notthe place for a discourse on the history of the concept. Suffice it to say that I have inmind the rather specific sense whose ontological foundation is an imagined separationbetween the human perceiver and the world, such that the perceiver has to reconstructthe world, in consciousness, prior to any meaningful engagement with it. The world ofnature, it is often said, is what lies ‘out there’. All kinds of entities are supposed to existout there, but not you and I. We live ‘in here’, in the intersubjective space marked outby our mental representations. Application of this logic forces an insistent dualism, betweenobject and subject, the material and the ideal, operational and cognised, ‘etic’ and ‘emic’.Some writers distinguish between nature and the landscape in just these terms – the formeris said to stand to the latter as physical reality to its cultural or symbolic construction.For example, Daniels and Cosgrove introduce a collection of essays on The iconographyof landscape with the following definition: ‘A landscape is a cultural image, a pictorial wayof representing or symbolising surroundings’ (1988: 1).

I do not share this view. To the contrary, I reject the division between inner and outerworlds – respectively of mind and matter, meaning and substance – upon which suchdistinction rests. The landscape, I hold, is not a picture in the imagination, surveyed bythe mind’s eye; nor however is it an alien and formless substrate awaiting the impositionof human order. ‘The idea of landscape’, as Meinig writes, ‘runs counter to recognitionof any simple binary relationship between man and nature’ (Meinig 1979b: 2). Thus,neither is the landscape identical to nature, nor is it on the side of humanity againstnature. As the familiar domain of our dwelling, it is with us, not against us, but it is noless real for that. And through living in it, the landscape becomes a part of us, just as weare a part of it. Moreover, what goes for its human component goes for other compo-nents as well. In a world construed as nature, every object is a self-contained entity,interacting with others through some kind of external contact. But in a landscape, eachcomponent enfolds within its essence the totality of its relations with each and every other.In short, whereas the order of nature is explicate, the order of the landscape is implicate(Bohm 1980: 172).

The landscape is not ‘space’. To appreciate the contrast, we could compare the everydayproject of dwelling in the world with the rather peculiar and specialised project of thesurveyor or cartographer whose objective is to represent it. No doubt the surveyor, as hegoes about his practical tasks, experiences the landscape much as does everyone else whosebusiness of life lies there. Like other people, he is mobile, yet unable to be in more thanone place at a time. In the landscape, the distance between two places, A and B, is expe-rienced as a journey made, a bodily movement from one place to the other, and thegradually changing vistas along the route. The surveyor’s job, however, is to take instru-mental measurements from a considerable number of locations, and to combine these datato produce a single picture which is independent of any point of observation. This pictureis of the world as it could be directly apprehended only by a consciousness capable ofbeing everywhere at once and nowhere in particular (the nearest we can get to this inpractice is by taking an aerial or bird’s-eye view). To such a consciousness, at once immo-bile and omnipresent, the distance between A and B would be the length of a line plottedbetween two points that are simultaneously in view, that line marking one of any number

123456789101231114567892012345678930123456789401234567118

The temporality of the landscape • 191 •

of journeys that could potentially be made (cf. Bourdieu 1977: 2). It is as though, froman imaginary position above the world, I could direct the movements of my body withinit, like a counter on a board, so that to say ‘I am here’ is not to point from somewhereto my surroundings, but to point from nowhere to the position on the board where mybody happens to be. And whereas actual journeys are made through a landscape, the boardon which all potential journeys may be plotted is equivalent to space.3

There is a tradition of geographical research (see, for example, Gould and White 1974)which sets out from the premise that we are all cartographers in our daily lives, and thatwe use our bodies as the surveyor uses his instruments, to register a sensory input frommultiple points of observation, which is then processed by our intelligence into an imagethat we carry around with us, like a map in our heads, wherever we go. The mind, ratherthan reaching into its surroundings from its dwelling place within the world, might belikened in this view to a film spread out upon its exterior surface. The sense of spaceimplicated in this cartographic view of environmental perception may be illuminated bymeans of an analogy drawn from the linguistics of Ferdinand de Saussure. To grasp theessence of language, Saussure invites us to picture thought and sound as two continuousand undifferentiated planes, of mental and phonic substance respectively, like two sidesof a sheet of paper. By cutting the sheet into pieces (words) we create, on one side, asystem of discrete concepts, and on the other, a system of discrete sounds; and since one side cannot be cut without at the same time cutting the other, the two systems ofdivision are necessarily homologous so that to each concept there corresponds a sound(Saussure 1959: 112–13).

Now when geographers and anthropologists write about space, what is generally impliedis something closely akin to Saussure’s sheet of paper, only in this case the counter-sideto thought is the continuum not of phonic substance but of the surface of the earth. Andso it appears that the division of the world into a mosaic of externally bounded segmentsis entailed in the very production of spatial meanings. Just as the word, for Saussure, isthe union of a concept with a delimited ‘chunk’ of sound, so the place is the union of asymbolic meaning with a delimited block of the earth’s surface. Spatial differentiationimplies spatial segmentation. This is not so of the landscape, however. For a place in thelandscape is not ‘cut out’ from the whole, either on the plane of ideas or on that ofmaterial substance. Rather, each place embodies the whole at a particular nexus within it,and in this respect is different from every other.

A place owes its character to the experiences it affords to those who spend time there– to the sights, sounds and indeed smells that constitute its specific ambience. And these,in turn, depend on the kinds of activities in which its inhabitants engage. It is from this relational context of people’s engagement with the world, in the business of dwelling,that each place draws its unique significance. Thus whereas with space, meanings areattached to the world, with the landscape they are gathered from it. Moreover, while placeshave centres – indeed it would be more appropriate to say that they are centres – they have no boundaries. In journeying from place A to place B it makes no sense to ask,along the way, whether one is ‘still’ in A or has ‘crossed over’ to B (Ingold 1986a: 155).Of course, boundaries of various kinds may be drawn in the landscape, and identifiedeither with natural features such as the course of a river or an escarpment, or with builtstructures such as walls and fences. But such boundaries are not a condition for the consti-tution of the places on either side of them; nor do they segment the landscape, for thefeatures with which they are identified are themselves an integral part of it. Finally, it isimportant to note that no feature of the landscape is, of itself, a boundary. It can only

123456789101234567892012345678930123456789401234567118

Dwelling• 192 •

become a boundary, or the indicator of a boundary, in relation to the activities of thepeople (or animals) for whom it is recognised or experienced as such.

In the course of explaining what the landscape is not, I have already moved some waytowards a positive characterisation. In short, the landscape is the world as it is known tothose who dwell therein, who inhabit its places and journey along the paths connectingthem. Is it not, then, identical to what we might otherwise call the environment? Certainlythe distinction between landscape and environment is not easy to draw, and for manypurposes they may be treated as practically synonymous. It will already be apparent thatI cannot accept the distinction offered by Yi-Fu Tuan, who argues that an environmentis ‘a given, a piece of reality that is simply there’, as opposed to the landscape, which isa product of human cognition, ‘an achievement of the mature mind’ (Tuan 1979: 90,100). For that is merely to reproduce the dichotomy between nature and humanity. The environment is no more ‘nature’ than is the landscape a symbolic construct. Elsewhere,I have contrasted nature and environment by way of a distinction between reality of –‘the physical world of neutral objects apparent only to the detached, indifferent observer’,and reality for – ‘the world constituted in relation to the organism or person whose environ-ment it is’ (Ingold 1992a: 44). But to think of environment in this sense is to regard itprimarily in terms of function, of what it affords to creatures – whether human or non-human – with certain capabilities and projects of action. Reciprocally, to regard thesecreatures as organisms is to view them in terms of their principles of dynamic functioning,that is as organised systems (Pittendrigh 1958: 394). As Lewontin succinctly puts it (1982:160), the environment is ‘nature organised by an organism’.

The concept of landscape, by contrast, puts the emphasis on form, in just the sameway that the concept of the body emphasises the form rather than the function of a livingcreature. If the body is the form in which a creature is present as a being-in-the-world,then the world of its being-in presents itself in the form of the landscape. Like organismand environment, body and landscape are complementary terms: each implies the other,alternately as figure and ground. The forms of the landscape are not, however, preparedin advance for creatures to occupy, any more than are the bodily forms of those creaturesindependently specified in their genetic make-up. Both sets of forms are generated andsustained in and through the processual unfolding of a total field of relations that cutsacross the emergent interface between organism and environment (Goodwin 1988). Havingregard to its formative properties, we may refer to this process as one of embodiment.

Though the notion of embodiment has recently come much into fashion, there hasbeen a tendency – following an ancient inclination in Western thought to prioritise formover process (Oyama 1985: 13) – to conceive of it as a movement of inscription, wherebysome pre-existing pattern, template or programme, whether genetic or cultural, is ‘realised’in a substantive medium. This is not what I have in mind, however. To the contrary,and adopting a helpful distinction from Paul Connerton (1989: 72–3), I regard embod-iment as a movement of incorporation rather than inscription, not a transcribing of formonto material but a movement wherein forms themselves are generated (Ingold 1990:215). Taking the organism as our focus of reference, this movement is what is commonlyknown as the life-cycle. Thus organisms may be said to incorporate, in their bodily forms, the life-cycle processes that give rise to them. Could not the same, then, be saidof the environment? Is it possible to identify a corresponding cycle, or rather a series of interlocking cycles, which builds itself into the forms of the landscape, and of whichthe landscape may accordingly be regarded as an embodiment? Before answering thisquestion, we need to turn to the second of my key terms, namely ‘temporality’.

123456789101231114567892012345678930123456789401234567118


TEMPORALITY

Let me begin, once again, by stating what temporality is not. It is not chronology (asopposed to history), and it is not history (as opposed to chronology). By chronology, I mean any regular system of dated time intervals, in which events are said to have takenplace. By history, I mean any series of events which may be dated in time according totheir occurrence in one or another chronological interval. Thus the Battle of Hastings wasan historical event, 1066 was a date (marking the interval of a year), and records tell usthat the former occurred in the latter. In the mere succession of dates there are no events,because everything repeats; in the mere succession of events there is no time, as nothingdoes. The relation between chronology and history, in this conception, has been wellexpressed by Kubler: ‘Without change there is no history; without regularity there is notime. Time and history are related as rule and variation: time is the regular setting forthe vagaries of history’ (1962: 72).

Now in introducing the concept of temporality, I do not intend that it should standas a third term, alongside the concepts of chronology and history. For in the sense inwhich I shall use the term here, temporality entails a perspective that contrasts radicallywith the one, outlined above, that sets up history and chronology in a relation of comple-mentary opposition. The contrast is essentially equivalent to that drawn by Alfred Gell(1992: 149–55) between what he calls (following McTaggart) the A-series, in which timeis immanent in the passage of events, and the B-series, in which events are strung out intime like beads on a thread. Whereas in the B-series, events are treated as isolated happen-ings, succeeding one another frame by frame, each event in the A-series is seen to encompassa pattern of retensions from the past and protentions for the future. Thus from the A-series point of view, temporality and historicity are not opposed but rather merge inthe experience of those who, in their activities, carry forward the process social life. Takentogether, these activities make up what I shall call the ‘taskscape’, and it is with the intrinsictemporality of the taskscape that I shall be principally concerned in this section.

We can make a start by returning for a moment to the distinction between land andlandscape. As a common denominator in terms of which constituents of the environmentof diverse kinds may be rendered quantitatively comparable, I compared land with weight.But I could equally have drawn the comparison with value or with labour. Value is thedenominator of commodities that enables us to say how much any one thing is worth bycomparison with another, even though these two things may be quite unlike in terms oftheir physical qualities and potential uses. In this sense, the concept of value (in general)is classically distinguished from that of use-value, which refers to the specific properties or‘affordances’ of any particular object, that commend it to the project of a user (Ingold1992a: 48–9, cf. Gibson 1979:127, Marx 1930: 169). Clearly, this distinction, betweenvalue and use-value, is precisely homologous to that between land and landscape. But ifwe turn to consider the work that goes into the making of useful things, then again wecan recognise that whilst the operations of making are indeed as unlike as the objectsproduced – involving different raw materials, different tools, different procedures anddifferent skills – they can nevertheless be compared in that they call for variable amountsof what may simply be called ‘labour’: the common denominator of productive activities.Like land and value, labour is quantitative and homogeneous, human work shorn of itsparticularities. It is of course the founding premise of the labour theory of value that theamount of value in a thing is determined by the amount of labour that went into producingit (I return to this theme in Chapter Seventeen, pp. 326–8).

123456789101234567892012345678930123456789401234567118

Dwelling• 194 •

How, then, should we describe the practices of work in their concrete particulars? Forthis purpose I shall adopt the term ‘task’, defined as any practical operation, carried outby a skilled agent in an environment, as part of his or her normal business of life. Inother words, tasks are the constitutive acts of dwelling. No more than features of the land-scape, however, are tasks suspended in a vacuum. Every task takes its meaning from itsposition within an ensemble of tasks, performed in series or in parallel, and usually bymany people working together. One of the great mistakes of recent anthropology – whatReynolds (1993: 410) calls ‘the great tool-use fallacy’ – has been to insist upon a separa-tion between the domains of technical and social activity, a separation that has blindedus to the fact that one of the outstanding features of human technical practices lies intheir embeddedness in the current of sociality. It is to the entire ensemble of tasks, intheir mutual interlocking, that I refer by the concept of taskscape. Just as the landscapeis an array of related features, so – by analogy – the taskscape is an array of related activ-ities. And as with the landscape, it is qualitative and heterogeneous: we can ask of ataskscape, as of a landscape, what it is like, but not how much of it there is. In short,the taskscape is to labour what the landscape is to land, and indeed what an ensemble ofuse-values is to value in general.

Now if value is measured out in units of money, and land in units of space, what isthe currency of labour? The answer, of course, is time – but it is time of a very peculiarsort, one that must be wholly indifferent to the modulations of human experience. Tomost of us it appears in the familiar guise of clock-time: thus an hour is an hour, regard-less of what one is doing in it, or of how one feels. But this kind of chronological timedoes not depend upon the existence of artificial clocks. It may be based on any perfectlyrepetitive, mechanical system, including that (putatively) constituted by the earth in itsaxial rotations and in its revolutions around the sun. Sorokin and Merton (1937), in aclassic paper, call it ‘astronomical’ time: it is, they write, ‘uniform, homogeneous; . . . purely quantitative, shorn of qualitative variations’. And they distinguish it from ‘socialtime’, which they see as fundamentally qualitative, something to which we can affix moral judgements such as good or bad, grounded in the ‘rhythms, pulsations and beatsof the societies in which they are found’, and for that reason tied to the particular circum-stances of place and people (1937: 621–3; see also Chapter Seventeen, pp. 325–6).Adopting Sorokin and Merton’s distinction, we could perhaps conclude that whereas labour is measured out in units of astronomical time, or in clock-time calibrated to anastronomical standard, the temporality of the taskscape is essentially social. Before we canaccept this conclusion, however, the idea of social time must be examined a little moreclosely.

In my earlier discussion of the significance of space, I showed that in the cartographicimagination, the mind is supposed to be laid out upon the surface of the earth. Likewisein the chronological perspective, time appears as the interface between mind and ‘duration’– by which is meant an undifferentiated stream of bodily activity and experience. Takingtime in this sense, Durkheim famously likened it to ‘an endless chart, where all durationis spread out before the mind, and upon which all possible events can be located in rela-tion to fixed and determinate guidelines’ (1976 [1915]: 10). Rather like Saussure’s sheetof paper, it could be compared to a strip of infinite length, with thought on one side andduration on the other. By cutting the strip into segments we establish a division, on theone hand, into calendrical intervals or dates, and on the other hand, into discrete ‘chunks’of lived experience, such that to every chunk there corresponds a date in a uniformsequence of before and after. And as every chunk succeeds the next, like frames on a reel

123456789101231114567892012345678930123456789401234567118


of film, we imagine ourselves to be looking on ‘as time goes by’, as though we could takeup a point of view detached from the temporal process of our life in the world and watchourselves engaged now in this task, now in that, in an unending series of present instants.Whence, then, come the divisions which give chronological form to the substance ofexperience? Durkheim’s answer, as is well known, was that these divisions – ‘indispens-able guidelines’ for the temporal ordering of events – come from society, correspondingto the ‘periodical recurrence of rites, feasts, and public ceremonies’ (p. 10). Thus forDurkheim, time is at once chronological and social, for society itself is a kind of clock,whose moving parts are individual human beings (Ingold 1986b: 341).

This is not, however, the way we perceive the temporality of the taskscape. For we doso not as spectators but as participants, in the very performance of our tasks. As Merleau-Ponty put it, in reckoning with an environment, I am ‘at my task rather than confrontingit’ (1962: 416). The notion that we can stand aside and observe the passage of time isfounded upon an illusion of disembodiment. This passage is, indeed, none other than our own journey through the taskscape in the business of dwelling. Once again we cantake our cue from Merleau-Ponty: ‘the passage of one present to the next is not a thingwhich I conceive, nor do I see it as an onlooker, I effect it’ (1962: 421). Reaching outinto the taskscape I perceive, at this moment, a particular vista of past and a future; butit is a vista that is available from this moment and no other (see Gell 1992: 269). Assuch, it constitutes my present, conferring upon it a unique character. Thus the present isnot marked off from a past that it has replaced or a future that will, in turn, replace it;it rather gathers the past and future into itself, like refractions in a crystal ball. And justas in the landscape, we can move from place to place without crossing any boundary,since the vista that constitutes the identity of a place changes even as we move, so like-wise can we move from one present to another without having to break through anychronological barrier that might be supposed to separate each present from the next inline. Indeed the features that Durkheim identified as serving this segmenting function –rites, feasts and ceremonies – are themselves as integral to the taskscape as are boundarymarkers such as walls or fences to the landscape.

The temporality of the taskscape is social, then, not because society provides an externalframe against which particular tasks find independent measure, but because people, in theperformance of their tasks, also attend to one another. Looking back, we can see thatDurkheim’s error was to divorce the sphere of people’s mutual involvement from that oftheir everyday practical activity in the world, leaving the latter to be carried on by indi-viduals in hermetic isolation. In real life, this is not how we go about our business. Bywatching, listening, perhaps even touching, we continually feel each other’s presence inthe social environment, at every moment adjusting our movements in response to thisongoing perceptual monitoring. For the orchestral musician, playing an instrument,watching the conductor and listening to one’s fellow players are all inseparable aspects ofthe same process of action: for this reason, the gestures of the performers may be said to resonate with each other. In orchestral music, the achievement of resonance – or whatSchutz (1951: 78) called a ‘mutual tuning-in relationship’ – is an absolute preconditionfor successful performance. But the same is true, more generally, of social life (Wikan1992, Richards 1996). Indeed it could be argued that in the resonance of movement andfeeling stemming from people’s mutually attentive engagement, in shared contexts of prac-tical activity, lies the very foundation of sociality.

Let me pursue the analogy between orchestral performance and social life a little furthersince, more than any other artistic genre, music mirrors the temporal form of the taskscape.

123456789101234567892012345678930123456789401234567118

Dwelling• 196 •

I want, by means of this analogy, to make three points. First, while there are cycles andrepetitions in music as in social life, these are essentially rhythmic rather than metronomic(on this distinction, see Young 1988: 19). It is for precisely this reason that social time,pace Durkheim, is not chronological. A metronome, like a clock, inscribes an artificialdivision into equal segments upon an otherwise undifferentiated movement; rhythm, bycontrast, is intrinsic to the movement itself. Langer has argued that the essence of rhythmlies in the successive building up and resolution of tension, on the principle that everyresolution is itself a preparation for the next building-up (1953: 126–7). There may ofcourse be rests or sustained notes within a piece, but far from breaking it up into segments,such moments are generally ones of high tension, whose resolution becomes ever moreurgent the longer they are held. Only our last exhalation of breath is not a preparationfor the next inhalation – with that, we die; similarly with the last beat the music comesto an end. Social life, however, is never finished, and there are no breaks in it that arenot integral to its tensile structure, to the ‘ebb and flow of activity’ by which society itselfseems to breathe (Young 1988: 53).

My second point is that in music as in social life, there is not just one rhythmic cycle,but a complex interweaving of very many concurrent cycles.4 While it reflects the temporalform of social life, music in fact represents a very considerable simplification, since itinvolves only one sensory register (the auditory), and its rhythms are fewer and moretightly controlled. In both cases, however, since any rhythm may be taken as the tempofor any of the others, there is no single, one-dimensional strand of time. As Langer putsit: ‘life is always a dense fabric of concurrent tensions, and as each of them is a measureof time, the measurements themselves do not coincide’ (1953: 113). Thus the temporalityof the taskscape, while it is intrinsic rather than externally imposed (metronomic), lies notin any particular rhythm, but in the network of interrelationships between the multiplerhythms of which the taskscape is itself constituted. To cite a celebrated anthropologicalexample: among the Nuer of southern Sudan, according to Evans-Pritchard, the passageof time is ‘primarily the succession of [pastoral] tasks and their relations to one another’(1940: 101–2, my emphasis). Each of these relations is, of course, a specific resonance.And so, just as social life consists in the unfolding of a field of relationships among personswho attend to one another in what they do, its temporality consists in the unfolding ofthe resultant pattern of resonances.

Thirdly, the forms of the taskscape, like those of music, come into being through move-ment. Music exists only when it is being performed; it does not pre-exist, as is sometimesthought, in the score, any more than a cake pre-exists in the recipe for making it. Similarly,the taskscape exists only so long as people are actually engaged in the activities of dwelling,despite the attempts of anthropologists to translate it into something rather equivalent toa score – a kind of ideal design for dwelling – that generally goes by the name of ‘culture’,and that people are supposed to bring with them into their encounter with the world.This parallel, however, brings me to a critical question. Up to now, my discussion oftemporality has concentrated exclusively on the taskscape, allowing the landscape to slipfrom view. It is now high time to bring it back into focus. I argued in the previous sectionthat the landscape is not nature; here I claim that the taskscape is not culture. Landscapeand taskscape, then, are not to be opposed as nature to culture. So how are we to under-stand the relation between them? Where does one end and the other begin? Can theyeven be distinguished at all?

If music best reflects the forms of the taskscape, it might be thought that painting isthe most natural medium for representing the forms of the landscape. And this suggests

123456789101231114567892012345678930123456789401234567118


that an examination of the difference, in the field of art, between music and paintingmight offer some clues as to how a distinction might possibly be drawn between taskscapeand landscape as facets of the real world. I begin by following up this suggestion.

TEMPORALISING THE LANDSCAPE

At first glance the difference seems obvious: paintings do not have to be performed, theyare presented to us as works that are complete in themselves. But on closer inspection,this contrast appears more as an artefact of a systematic bias in Western thought, to whichI have already alluded, that leads us to privilege form over process. Thus the actual workof painting is subordinated to the final product; the former is hidden from view so thatthe latter alone becomes an object of contemplation. In many non-Western societies, bycontrast, the order of priority is reversed: what is essential is the act of painting itself, ofwhich the products may be relatively short-lived – barely perceived before being erasedor covered up. This is so, for example, among the Yolngu, an Aboriginal people of northernAustralia, whose experience of finished paintings, according to their ethnographer, is limitedto ‘images fleetingly glimpsed out of the corner of the eye’ (Morphy 1992: 187). Theemphasis, here, is on painting as performance.5 Far from being the preparation of objectsfor future contemplation, it is an act of contemplation in itself. So, too, is performing orlistening to music. Thus all at once, the contrast between painting and music seems lesssecure. It becomes a matter of degree, in the extent to which forms endure beyond theimmediate contexts of their production. Musical sound, of course, is subject to the prop-erty of rapid fading: speeding outwards from its point of emission, and dissipating as itgoes, it is present only momentarily to our senses. But where, as in painting, gesturesleave their traces in solid substance, the resulting forms may last much longer, albeit neverindefinitely.

Returning now from the contrast between music and painting to that between taskscapeand landscape, the first point to note is that no more than a painting is the landscapegiven ready-made. One cannot, as Inglis points out, ‘treat landscape as an object if it isto be understood. It is a living process; it makes men; it is made by them’ (1977: 489).Just as with music, the forms of the landscape are generated in movement: these forms,however, are congealed in a solid medium – indeed, to borrow Inglis’s words again, ‘a landscape is the most solid appearance in which a history can declare itself ’ (ibid.).Thanks to their solidity, features of the landscape remain available for inspection longafter the movement that gave rise to them has ceased. If, as Mead argued (1977 [1938]:97), every object is to be regarded as a ‘collapsed act’, then the landscape as a whole mustlikewise be understood as the taskscape in its embodied form: a pattern of activities ‘collapsed’into an array of features.

But to reiterate a point made earlier, the landscape takes on its forms through a processof incorporation, not of inscription. That is to say, landscape formation is not a matter– as James Weiner would have it (1991: 32) – of transforming ‘a sheer physical terraininto a pattern of historically experienced and constituted space and time’, as though thephysical world pre-existed as a blank slate, a mere substrate of formless materiality, awaitingthe impress of cultural significance. Human beings do not, in their movements, inscribetheir life histories upon the surface of nature as do writers upon the page; rather, thesehistories are woven, along with the life-cycles of plants and animals, into the texture ofthe surface itself (see Chapter Eighteen, pp. 347–8). Thus the forms of the landscape arisealongside those of the taskscape, within the same current of activity. If we recognise a

123456789101234567892012345678930123456789401234567118

Dwelling• 198 •

man’s gait in the pattern of his footprints, it is not because the gait preceded the foot-prints and was ‘inscribed’ in them, but because both the gait and the prints arose withinthe movement of the man’s walking.

Since, moreover, the activities that comprise the taskscape are unending, the landscapeis never complete: neither ‘built’ nor ‘unbuilt’, it is perpetually under construction. Thisis why the conventional dichotomy between natural and artificial (or ‘man-made’) compo-nents of the landscape is so problematic. Virtually by definition, an artefact is an objectshaped to a pre-conceived image that motivated its construction, and it is finished at thepoint when it is brought into conformity with this image. What happens to it beyondthat point is supposed to belong to the phase of use rather than manufacture, to dwellingrather than building. But the forms of the landscape are not pre-prepared for people tolive in – not by nature nor by human hands – for it is in the very process of dwellingthat these forms are constituted. We may recall here Heidegger’s remark, already cited in the last chapter, that ‘to build is in itself already to dwell’ (1971: 146). Thus the landscape is always in the nature of work in progress.

My conclusion that the landscape is the congealed form of the taskscape does enableus to explain why, intuitively, the landscape seems to be what we see around us, whereasthe taskscape is what we hear. To be seen, a thing need do nothing itself, for the opticarray that specifies its form to a viewer consists of light reflected off its outer surfaces. To be heard, on the other hand, a thing must actively emit sounds or, through its move-ment, cause sound to be emitted by other objects with which it comes into contact. Thus,outside my window I see a landscape of houses, trees, gardens, a street and pavement. Ido not hear any of these things, but I can hear people talking on the pavement, a carpassing by, birds singing in the trees, a dog barking somewhere in the distance, and thesound of hammering as a neighbour repairs his garden shed. In short, what I hear isactivity, even when its source cannot be seen. And since the forms of the taskscape,suspended as they are in movement, are present only as activity, the limits of the taskscapeare also the limits of the auditory world. (While I deal here only with visual and auralperception, we should not underestimate the significance of touch, which is important toall of us but above all to blind people, for whom it opens up the possibility of access to the landscape – if only through proximate bodily contact.)

This argument carries an important corollary. Whereas both the landscape and thetaskscape presuppose the presence of an agent who watches and listens, the taskscape mustbe populated with beings who are themselves agents, and who reciprocally ‘act back’ inthe process of their own dwelling. In other words, the taskscape exists not just as activitybut as interactivity. Indeed this conclusion was already foreshadowed when I introducedthe concept of resonance as the rhythmic harmonisation of mutual attention. Having saidthat, however, there is no reason why the domain of interactivity should be confined tothe movements of human beings. We hear animals as well as people, such as the birdsand the dog in my example above. Hunters, to take another example, are alert to everysight, sound or smell that reveals the presence of animals, and we can be sure that theanimals are likewise alert to the presence of humans, as they are also to that of one another.On a larger scale, the hunters’ journeys through the landscape, or their oscillations betweenthe procurement of different animal species, resonate with the migratory movements ofterrestrial mammals, birds and fish. Perhaps then, as Reed argues, there is a fundamentaldifference between our perception of animate beings and inanimate objects, since theformer – by virtue of their capacity for autonomous movement – ‘are aware of theirsurroundings (including us) and because they act on those surroundings (including us)’

123456789101231114567892012345678930123456789401234567118


(Reed 1988a: 116). In other words, they afford the possibility not only of action but alsoof interaction (cf. Gibson 1979: 135). Should we, then, draw the boundaries of thetaskscape around the limits of the animate?

Though the argument is a compelling one I find it unsatisfactory, for two reasons inparticular. First, as Langer observes, ‘rhythm is the basis of life, but not limited to life’(1953: 128). The rhythms of human activities resonate not only with those of other livingthings but also with a whole host of other rhythmic phenomena – the cycles of day andnight and of the seasons, the winds, the tides, and so on. Citing a petition of 1800 fromthe seaside town of Sunderland, in which it is explained that ‘people are obliged to beup at all hours of the night to attend the tides and their affairs upon the river’, Thompson(1967: 59–60) notes that ‘the operative phrase is “attend the tides”: the patterning ofsocial time in the seaport follows upon the rhythms of the sea’. In many cases these naturalrhythmic phenomena find their ultimate cause in the mechanics of planetary motion, but it is not of course to these that we resonate. Thus we resonate to the cycles of lightand darkness, not to the rotation of the earth, even though the diurnal cycle is caused bythe earth’s axial rotation. And we resonate to the cycles of vegetative growth and decay,not to the earth’s revolutions around the sun, even though the latter cause the cycle ofthe seasons. Moreover these resonances are embodied, in the sense that they are not onlyhistorically incorporated into the enduring features of the landscape but also develop-mentally incorporated into our very constitution as biological organisms. Thus Youngdescribes the body as ‘an array of interlocking (or interflowing) cycles, with their ownspheres of partial independence within the solar cycle’ (1988: 41). We do not consultthese cycles, as we might consult a wrist-watch, in order to time our own activities, forthe cycles are inherent in the rhythmic structure of the activities themselves. It wouldseem, then, that the pattern of resonances that comprises the temporality of the taskscapemust be expanded to embrace the totality of rhythmic phenomena, whether animate orinanimate.

The second reason why I would be reluctant to restrict the taskscape to the realm ofliving things has to do with the very notion of animacy. I do not think we can regardthis as a property that can be ascribed to objects in isolation, such that some (animate)have it and others (inanimate) do not. For life is not a principle that is separately installedinside individual organisms, and which sets them in motion upon the stage of the inan-imate. To the contrary, as I have argued elsewhere, life is ‘a name for what is going on inthe generative field within which organic forms are located and “held in place” ’ (Ingold1990: 215). That generative field is constituted by the totality of organism–environmentrelations, and the activities of organisms are moments of its unfolding. Indeed once wethink of the world in this way, as a total movement of becoming which builds itself intothe forms we see, and in which each form takes shape in continuous relation to thosearound it, then the distinction between the animate and the inanimate seems to dissolve.The world itself takes on the character of an organism, and the movements of animals –including those of us human beings – are parts or aspects of its life-process (Lovelock1979). This means that in dwelling in the world, we do not act upon it, or do things to it; rather we move along with it. Our actions do not transform the world, they are partand parcel of the world’s transforming itself. And that is just another way of saying thatthey belong to time.

For in the final analysis, everything is suspended in movement. David Reason expressesthe point in an eloquent passage that could stand as a summary of all I have argued so far:

123456789101234567892012345678930123456789401234567118

Dwelling• 200 •

Landscapes change; and change is itself an intrinsic aspect of our experience of land-scape. The landscape is a polyrhythmic composition of processes whose pulse variesfrom the erratic flutter of leaves to the measured drift and clash of tectonic plates.Relative to the human span, the view before us seems composed of fleeting, ephemeraleffects which create a patina of transience on apparently stable forms.

(1987: 40)

As this passage reveals, what appear to us as the fixed forms of the landscape, passive andunchanging unless acted upon from outside, are themselves in motion, albeit on a scaleimmeasurably slower and more majestic than that on which our own activities areconducted. Imagine a film of the landscape, shot over years, centuries, even millennia.Slightly speeded up, plants appear to engage in very animal-like movements, trees flextheir limbs without any prompting from the winds. Speeded up rather more, glaciers flowlike rivers and even the earth begins to move. At yet greater speeds solid rock bends,buckles and flows like molten metal. The world itself begins to breathe. Thus the rhythmicpattern of human activities nests within the wider pattern of activity for all animal life,which in turn nests within the pattern of activity for all so-called living things, whichnests within the life-process of the world.

At each of these levels, as Mae-Wan Ho shows, coherence is founded upon resonance(Ho 1989: 18–20). Reminding us of Whitehead’s maxim, that there is ‘no holding naturestill and looking at it’, Ho argues that the world is not anything we can look at but aprocess that we are part of. Ultimately, then, by re-placing the tasks of human dwellingin their proper context within the process of becoming of the world as a whole, we cando away with the dichotomy between taskscape and landscape – only, however, by recog-nising the fundamental temporality of the landscape itself.6

THE HARVESTERS

In order to provide some illustration of the ideas developed in the preceding sections, Ireproduce here a painting which, more than any other I know, vividly captures a senseof the temporality of the landscape. This is The harvesters, painted by Pieter Bruegel theElder in 1565 (Figure 11.1). I am not an art historian or critic, and my purpose is notto analyse the painting in terms of style, composition or aesthetic effect. Nor am Iconcerned with the historical context of its production. Suffice it to say that the pictureis believed to be one of a series of twelve, each depicting a month of the year, out ofwhich only five have survived (Gibson 1977: 147). Each panel portrays a landscape, inthe colours and apparel appropriate to the month, and shows people engaged in the tasksof the agricultural cycle that are usual at that time of year. The harvesters depicts themonth of August, and shows field hands at work, reaping and sheafing a luxuriant cropof wheat, whilst others pause for a midday meal and some well-earned rest. The sense ofrustic harmony conveyed in this scene may, perhaps, represent something of an idealisa-tion on Bruegel’s part. As Walter Gibson points out, Bruegel was inclined to ‘depictpeasants very much as a wealthy landowner would have viewed them, as the anonymoustenders of his fields and flocks’ (1977: 157–8). Any landowner would have had cause forsatisfaction in such a fine crop, whereas the hands who sweated to bring it in may havehad a rather different experience. Nevertheless, Bruegel painted during a period of greatmaterial prosperity in the Netherlands, in which all shared to some degree. These werefortunate times.

123456789101231114567892012345678930123456789401234567118


We are accustomed, by the conventions of modern society, to describe our experienceof landscape as though we were viewing a picture. What I am about to suggest, however,is precisely the reverse. Rather than treating the world as its own painting I should likeyou, the reader, to regard this painting by Bruegel as though it were its own world, intowhich you have been magically transported. Imagine yourself, then, set down in the verylandscape depicted, on a sultry August day in 1565. Standing a little way off to the rightof the group beneath the tree, you are a witness to the scene unfolding about you. Andof course you hear it too, for the scene does not unfold in silence. So used are we tothinking of the landscape as a picture that we can look at, like a plate in a book or animage on a screen, that it is perhaps necessary to remind you that exchanging the paintingfor ‘real life’ is not simply a matter of increasing the scale. What is involved is a funda-mental difference of orientation. In the landscape of our dwelling, we look around (Gibson1979: 203).7 In what follows I shall focus on six components of what you see aroundyou, and comment on each insofar as they illustrate aspects of what I have had to sayabout landscape and temporality. They are: the hills and valley, the paths and tracks, thetree, the corn, the church, and the people.

123456789101234567892012345678930123456789401234567118

Dwelling• 202 •

Figure 11.1 The harvesters (1565) by Pieter Bruegel the Elder.

The Metropolitan Museum of Art, Rogers Fund, 1919 (19.164).

The hills and valley

The terrain is a gently undulating one of low hills and valleys, grading off to a shorelinethat can just be made out through the summer haze. You are standing near the summitof a hill, from where you can look out across the intervening valley to the next. How,then, do you differentiate between the hills and the valley as components of this land-scape? Are they alternating blocks or strips into which it may be divided up? Any attemptat such division plunges us immediately into absurdity. For where can we draw the bound-aries of a hill except along the valley bottoms that separate it from the hills on either side?And where can we draw the boundaries of a valley except along the summits of the hillsthat mark its watershed? One way, we would have a landscape consisting only of hills,the other way it would consist only of valleys. Of course, ‘hill’ and ‘valley’ are opposedterms, but the opposition is not spatial or altitudinal but kinaesthetic. It is the move-ments of falling away from, and rising up towards, that specify the form of the hill; andthe movements of falling away towards, and rising up from, that specify the form of thevalley. Through the exercises of descending and climbing, and their different muscularentailments, the contours of the landscape are not so much measured as felt – they aredirectly incorporated into our bodily experience. But even if you remain rooted to onespot, the same principle applies. As you look across the valley to the hill on the horizon,your eyes do not remain fixed: swivelling in their sockets, or as you tilt your head, theirmotions accord with the movement of your attention as it follows its course through thelandscape. You cast your eyes first downwards into the valley, and then upwards towardsthe distant hill. Indeed in this vernacular phrase, to ‘cast one’s eyes’, common sense hasonce again grasped intuitively what the psychology of vision, with its metaphors of retinalimagery, has found so hard to accept: that movement is the very essence of perception.It is because, in scanning the terrain from nearby into the distance, your downward glanceis followed by an upward one, that you perceive the valley.

Moreover someone standing where you are now would perceive the same topographicpanorama, regardless of the time of year, the weather conditions and the activities in whichpeople may be engaged. We may reasonably suppose that over the centuries, perhaps evenmillennia, this basic topography has changed but little. Set against the duration of humanmemory and experience, it may therefore be taken to establish a baseline of permanence.Yet permanence, as Gibson has stressed, is always relative; thus ‘it is better to speak ofpersistence under change’ (Gibson 1979: 13). Although the topography is invariant rela-tive to the human life-cycle, it is not itself immune to change. Sea-levels rise and fall withglobal climatic cycles, and the present contours of the country are the cumulative outcomeof a slow and long drawn out process of erosion and deposition. This process, moreover,was not confined to earlier geological epochs during which the landscape assumed itspresent topographic form. For it is still going on, and will continue so long as the stream,just visible in the valley bottom, flows on towards the sea. The stream does not flowbetween pre-cut banks, but cuts its banks even as it flows. Likewise, as we have seen,people shape the landscape even as they dwell. And human activities, as well as the actionof rivers and the sea, contribute significantly to the process of erosion. As you watch, thestream flows, folk are at work, a landscape is being formed, and time passes.

The paths and tracks

I remarked above that we experience the contours of the landscape by moving throughit, so that it enters – as Bachelard would say – into our ‘muscular consciousness’. Reliving

123456789101231114567892012345678930123456789401234567118


the experience in our imagination, we are inclined to recall the road we took as ‘climbing’the hill, or as ‘descending’ into the valley, as though ‘the road itself had muscles, or rather,counter-muscles’ (Bachelard 1964: 11). And this, too, is probably how you recall the pathsand tracks that are visible to you now: after all, you must have travelled along at leastsome of them to reach the spot where you are currently standing. Nearest at hand, a pathhas been cut through the wheat-field, allowing sheaves to be carried down, and water andprovisions to be carried up. Further off, a cart-track runs along the valley bottom, andanother winds up the hill behind. In the distance, paths criss-cross the village green. Takentogether, these paths and tracks ‘impose a habitual pattern on the movement of people’(Jackson 1989: 146). And yet they also arise out of that movement, for every path ortrack shows up as the accumulated imprint of countless journeys that people have made– with or without their vehicles or domestic animals – as they have gone about theireveryday business. Thus the same movement is embodied, on the side of the people, intheir ‘muscular consciousness’, and on the side of the landscape, in its network of pathsand tracks. In this network is sedimented the activity of an entire community, over manygenerations. It is the taskscape made visible.

In their journeys along paths and tracks, however, people also move from place to place.To reach a place, you need cross no boundary, but you must follow some kind of path.Thus there can be no places without paths, along which people arrive and depart; andno paths without places, that constitute their destinations and points of departure. Andfor the harvesters, the place to which they arrive, and whence they will leave at the endof the day, is marked by the next feature of the landscape to occupy your attention . . .

The tree

Rising from the spot where people are gathered for their repast is an old and gnarledpear-tree, which provides them with both shade from the sun, a back-rest and a prop forutensils. Being the month of August, the tree is in full leaf, and fruit is ripening on thebranches. But this is not just any tree. For one thing, it draws the entire landscape aroundit into a unique focus: in other words, by its presence it constitutes a particular place.The place was not there before the tree, but came into being with it. And for those whoare gathered there, the prospect it affords, which is to be had nowhere else, is what givesit its particular character and identity. For another thing, no other tree has quite the sameconfiguration of branches, diverging, bending and twisting in exactly the same way. In itspresent form, the tree embodies the entire history of its development from the momentit first took root. And that history consists in the unfolding of its relations with mani-fold components of its environment, including the people who have nurtured it, tilledthe soil around it, pruned its branches, picked its fruit, and – as at present – use it assomething to lean against. The people, in other words, are as much bound up in the lifeof the tree as is the tree in the lives of the people.8 Moreover, unlike the hills and thevalley, the tree has manifestly grown within living memory. Thus its temporality is moreconsonant with that of human dwelling. Yet in its branching structure, the tree combinesan entire hierarchy of temporal rhythms, ranging from the long cycle of its own germi-nation, growth and eventual decay to the short, annual cycle of flowering, fruiting andfoliation. At one extreme, represented by the solid trunk, it presides immobile over thepassage of human generations; at the other, represented by the frondescent shoots, itresonates with the life-cycles of insects, the seasonal migrations of birds, and the regularround of human agricultural activities (Davies 1988). In a sense, then, the tree bridges

123456789101234567892012345678930123456789401234567118

Dwelling• 204 •

the gap between the apparently fixed and invariant forms of the landscape and the mobileand transient forms of animal life, visible proof that all of these forms, from the mostpermanent to the most ephemeral, are dynamically linked under transformation withinthe movement of becoming of the world as a whole.

The corn

Turning from the pear-tree to the wheat-field, it is no longer a place in the landscape butthe surrounding surface that occupies your attention. And perhaps what is most strikingabout this surface is its uniformity of colour, a golden sheen that cloaks the more elevatedparts of the country for as far as the eye can see. As you know, wheat takes on this colourat the particular time of year when it is ripe for harvesting. More than any other featureof the landscape, the golden corn gathers the lives of its inhabitants, wherever they maybe, into temporal unison, founded upon a communion of visual experience. Thus whereasthe tree binds past, present and future in a single place, the corn binds every place in thelandscape within a single horizon of the present. The tree, we could say, establishes avivid sense of duration, the corn an equally vivid sense of what Fabian (1983: 31) callscoevalness. It is this distinction that Bachelard has in mind when he contrasts the ‘before-me, before-us’ of the forest with the ‘with-me, with-us’ of fields and meadows, wherein‘my dreams and recollections accompany all the different phases of tilling and harvesting’(Bachelard 1964: 188). You may suppose that the sleeper beneath the tree is dreaming ofcorn, but if so, you may be sure that the people and the activities that figure in his dreamare coeval with those of the present and do not take him back into an encounter withthe past.9

Where the corn has been freshly cut, it presents a sheer vertical front, not far short ofa man’s height. But this is not a boundary feature, like a hedge or fence. It is an inter-face, whose outline is progressively transformed as the harvesters proceed with their work.Here is a fine example of the way in which form emerges through movement. Anotherexample can be seen further off, where a man is engaged in the task of binding the wheat into a sheaf. Each completed sheaf has a regular form, which arises out of the coordinated movement of binding. But the completion of a sheaf is only one moment in the labour process. The sheaves will later be carried down the path through the field,to the haycart in the valley. Indeed at this very moment, one woman is stooped almostdouble in the act of picking up a sheaf, and two others can be seen on their way down,sheaves on their shoulders. Many more operations will follow before the wheat is even-tually transformed into bread. In the scene before you, one of the harvesters under thetree, seated on a sheaf, is cutting a loaf. Here the cycle of production and consumptionends where it began, with the producers. For production is tantamount to dwelling: itdoes not begin here (with a pre-conceived image) and end there (with a finished artefact),but is continuously going on.

The church

Not far off, nestled in a grove of trees near the top of the hill, is a stone church. It isinstructive to ask: how does the church differ from the tree? They have more in common,perhaps, than meets the eye. Both possess the attributes of what Bakhtin (1981: 84) callsa ‘chronotope’ – that is, a place charged with temporality, one in which temporality takeson palpable form. Like the tree, the church by its very presence constitutes a place, which

123456789101231114567892012345678930123456789401234567118


owes its character to the unique way in which it draws in the surrounding landscape.Again like the tree, the church spans human generations, yet its temporality is not incon-sonant with that of human dwelling. As the tree buries its roots in the ground, so alsopeople’s ancestors are buried in the graveyard beside the church, and both sets of rootsmay reach to approximately the same temporal depth. Moreover the church, too, resonatesto the cycles of human life and subsistence. Among the inhabitants of the neighbourhood,it is not only seen but also heard, as its bells ring out the seasons, the months, births,marriages and deaths. In short, as features of the landscape, both the church and the treeappear as veritable monuments to the passage of time.

Yet despite these similarities, the difference may seem obvious. The church, after all, isa building. The tree, by contrast, is not built, it grows. We may agree to reserve the term‘building’ for any durable structure in the landscape whose form arises and is sustainedwithin the current of human activity. It would be wrong to conclude, however, that thedistinction between buildings and non-buildings is an absolute one. Where an absolutedistinction is made, it is generally founded on the assumption that built form, rather thanhaving its source within nature, is superimposed by the mind upon it. That assumption,however, presupposes the separation of mind and nature. But from the perspective ofdwelling there is no such separation. It is evident, from this latter perspective, that theforms of buildings, as much as of any other features of the landscape, are neither givenin the world nor placed upon it, but emerge within the self-transforming processes of theworld itself. With respect to any feature, the scope of human involvement in these processeswill vary from negligible to considerable, though it is never total (even the most engi-neered of environments is home to other species). Thus to recall our conclusion from thelast chapter, what is or is not a building is a relative matter; moreover as human involve-ment may vary in the life history of a feature, it may be more or less of a building indifferent periods.

Returning to the tree and the church, it is clearly too simple to suppose that the formof the tree is naturally given in its genetic make-up, whereas the form of the church pre-exists, in the minds of the builders, as a plan which is then realised in stone. In the caseof the tree, we have already observed that its growth consists in the unfolding of a totalsystem of relations constituted by the fact of its presence in an environment, from thepoint of germination onwards, and that people, as components of the tree’s environment,play a not insignificant role in this process. Likewise the ‘biography’ of the church consistsin the unfolding of relations with its human builders, as well as with other componentsof its environment, from the moment when the first stone was laid. The final form ofthe church may indeed have been prefigured in the human imagination, but it no moreissued from the image than did the form of the tree issue from its genes. In both cases,the form is the embodiment of a developmental or historical process, and is rooted in thecontext of human dwelling in the world.

In the case of the church, moreover, that process did not stop when its form came tomatch the conceptual model. For as long as the building remains standing in the land-scape, it will continue – as it does now – to figure within the environment not just ofhuman beings but of a myriad of other living kinds, plant and animal, which will incor-porate it into their own life-activities and modify it in the process. And it is subject, too,to the same forces of weathering and decomposition, both organic and meteorological,that affect everything else in the landscape. The preservation of the church in its existing,‘finished’ form in the face of these forces, however substantial it may be in its materialsand construction, requires a regular input of effort in maintenance and repair. Once this

123456789101234567892012345678930123456789401234567118

Dwelling• 206 •

human input lapses, leaving it at the mercy of other forms of life and of the weather, itwill soon cease to be a building and become a ruin.

The people

So far I have described the scene only as you behold it with your eyes. Yet you do notonly look, you listen as well, for the air is full of sounds of one kind and another. Thoughthe folk beneath the tree are too busy eating to talk, you hear the clatter of wooden spoonson bowls, the slurp of the drinker, and the loud snores of the member of the party whois outstretched in sleep. Further off, you hear the swish of scythes against the cornstalksand the calls of the birds as they swoop low over the field in search of prey. Far off inthe distance, wafted on the light wind, can be heard the sounds of people conversing andplaying on a green, behind which, on the other side of the stream, lies a cluster of cottages.What you hear is a taskscape.

In the performance of their particular tasks, people are responsive not only to the cycleof maturation of the crop, which draws them together in the overall project of harvesting,but also to each other’s activities as these are apportioned by the division of labour. Evenwithin the same task, individuals do not carry on in mutual isolation. Technically, it takesonly one man to wield a scythe, but the reapers nevertheless work in unison, achieving adance-like harmony in their rhythmic movements. Similarly the two women carryingsheaves down into the valley adjust their pace, each in relation to the other, so that thedistance between them remains more or less invariant. Perhaps there is less co-ordinationbetween the respective movements of the eaters, however they eye each other intently asthey set about their repast, and the meal is a joint activity on which all have embarkedtogether, and which they will finish together. Only the sleeper, oblivious to the world, isout of joint – his snores jar the senses precisely because they are not in any kind ofrhythmic relation to what is going on around. Without wakeful attention, there can beno resonance.

But in attending to one another, do the people inhabit a world of their own, an exclu-sively human world of meanings and intentions, of beliefs and values, detached from theone in which their bodies are put to work in their several activities? Do they, from withinsuch a domain of intersubjectivity, look at the world outside through the window of theirsenses? Surely not. For the hills and valley, the tree, the corn and the birds are as palpablypresent to them (as indeed to you too) as are the people to each other (and to you). Thereapers, as they wield their scythes, are with the corn, just as the eaters are with theirfellows. The landscape, in short, is not a totality that you or anyone else can look at, itis rather the world in which we stand in taking up a point of view on our surroundings.And it is within the context of this attentive involvement in the landscape that the humanimagination gets to work in fashioning ideas about it. For the landscape, to recall thewords of Merleau-Ponty (1962: 24), is not so much the object as ‘the homeland of ourthoughts’.

EPILOGUE

Concluding an essay on the ways in which the Western Apache of Arizona discovermeaning, value and moral guidance in the landscape around them, Keith Basso abhorsthe tendency in ecological anthropology to relegate such matters to an ‘epiphenomenal’level, which is seen to have little or no bearing on the dynamics of adaptation of human

123456789101231114567892012345678930123456789401234567118


populations to the conditions of their environments. An ecology that is fully cultural,Basso argues, is one that would attend as much to the semiotic as to the material dimen-sions of people’s relations with their surroundings, by bringing into focus ‘the layers ofsignificance with which human beings blanket the environment’ (Basso 1984: 49). Inrather similar vein, Denis Cosgrove regrets the tendency in human geography to regardthe landscape in narrowly utilitarian and functional terms, as ‘an impersonal expressionof demographic and economic forces’, and thus to ignore the multiple layers of symbolicmeaning or cultural representation that are deposited upon it. The task of decoding the‘many-layered meanings of symbolic landscapes’, Cosgrove argues, will require a geographythat is not just human but properly humanistic (Cosgrove 1989: 120–7).

Though I have some sympathy with the views expressed by these writers, I believe thatthe metaphors of cultural construction they adopt have an effect quite opposite to thatintended. For the very idea that meaning covers over the world, layer upon layer, carriesthe implication that the way to uncover the most basic level of human beings’ practicalinvolvement with their environments is by stripping these layers away. In other words,such blanketing metaphors actually serve to create and perpetuate an intellectual space inwhich human ecology or human geography can flourish, untroubled by any concerns aboutwhat the world means to the people who live in it. We can surely learn from the WesternApache, who insist that the stories they tell, far from putting meanings upon the land-scape, are intended to allow listeners to place themselves in relation to specific features ofthe landscape, in such a way that their meanings may be revealed or disclosed. Storieshelp to open up the world, not to cloak it.

And such opening up, too, must be the objective of archaeology. Like the WesternApache – and for that matter any other group of people who are truly ‘at home’ in theworld – archaeologists study the meaning of the landscape not by interpreting the manylayers of its representation (adding further layers in the process) but by probing ever moredeeply into it. Meaning is there to be discovered in the landscape, if only we know howto attend to it. Every feature, then, is a potential clue, a key to meaning rather than avehicle for carrying it.10 This discovery procedure, wherein objects in the landscape becomeclues to meaning, is what distinguishes the perspective of dwelling. And since, as I haveshown, the process of dwelling is fundamentally temporal, the apprehension of the land-scape in the dwelling perspective must begin from a recognition of its temporality. Onlythrough such recognition, by temporalising the landscape, can we move beyond the divi-sion that has afflicted most inquiries up to now, between the scientific study of anatemporalised nature, and the humanistic study of a dematerialised history. And no disci-pline is better placed to take this step than archaeology. I have not been concerned herewith either the methods or the results of archaeological inquiry. However to the question,‘what is archaeology the study of ?’, I believe there is no better answer than ‘the tempo-rality of the landscape’. I hope, in this chapter, to have gone some way towards elucidatingwhat this means.

123456789101234567892012345678930123456789401234567118

Dwelling• 208 •

Chapter Twelve

Globes and spheresThe topology of environmentalism

My purpose in this chapter is no more than to try out a rather embryonic idea. It concernsthe significance of the image of the globe in the language of contemporary debate aboutthe environment. Though the image has long been deployed in geopolitical contexts, andeven longer in connection with navigation and astronomy, my impression is that its useas a characterisation of the environment is rather recent. I have in mind such phrases,which slip so readily off the tongues of contemporary policy-makers, as ‘global environ-mental change’. One is immediately struck by the paradoxical nature of this phrase. Anenvironment, surely, is that which surrounds, and can exist, therefore, only in relation towhat is surrounded (Ingold 1992a: 40). I do not think that those who speak of the globalenvironment mean by this the environment surrounding the globe. It is our environmentthey are talking about, the world as it presents itself to a universal humanity. Yet howcan humans, or for that matter beings of any other kind, possibly be surrounded by aglobe? Would it not be fairer to say that it is we who have surrounded it?

My idea is that what may be called the global outlook may tell us something importantabout the modern conception of the environment as a world which, far from being theambience of our dwelling, is turned in upon itself, so that we who once stood at its centrebecome first circumferential and are finally expelled from it altogether (Figure 12.1). Inother words, I am suggesting that the notion of the global environment, far from markinghumanity’s reintegration into the world, signals the culmination of a process of separation.

The image of the globe is familiarto all of us who have gone through aWestern schooling and are used tostudying models upon which aredrawn, in outline, the continents andoceans, and the gridlines of latitudeand longitude. We are taught thatthis is what the earth looks like,although none of us, with a handfulof significant exceptions, has ever seenit. By and large, life is lived at suchclose proximity to the earth’s surfacethat a global perspective is unobtain-able. The significant exceptionscomprise, of course, that privilegedband of astronauts who have viewedthe earth from outer space. In a sense,

123456789101231114567892012345678930123456789401234567118


A B

Figure 12.1 Two views of the environment: (A) as a lifeworld; (B)as a globe.

the astronaut’s relation to the real globeseen through the window of the space-craft mirrors the schoolchild’s relation tothe model globe in the classroom: inboth cases the world appears as an objectof contemplation, detached from thedomain of lived experience. For the childthe world is separately encapsulated inthe model; for the astronaut life is sepa-rately encapsulated, albeit temporarily, inthe space module. My point with thiscomparison is a simple one: with theworld imaged as a globe, far from cominginto being in and through a life process,it figures as an entity that is, as it were,presented to or confronted by life. Theglobal environment is not a lifeworld, itis a world apart from life.

Before pursuing the implications ofthis view, I should like to introduce analternative image of the world which, at least in European thought, is of farmore ancient provenance. This is theimage of the sphere. Something of the difference in connotation between

‘globe’ and ‘sphere’ is suggested in their very acoustic resonance: ‘globe’ is hard and conso-nantal; ‘sphere’ soft and vocalic. A globe is solid and opaque, a sphere hollow andtransparent. For the early astronomers, of course, the cosmos itself was seen to be comprisedof a series of such spheres, at the common centre of which stood man himself. The ideawas that as man’s attention was drawn ever outward, so it would penetrate each sphereso as to reach the next. This is illustrated in Figure 12.2, taken from the Scala Naturaleof Giovanni Camillo Maffei, published in Venice in 1564, and dedicated to the Countof Altavilla. Here there are fourteen concentric spheres which – Maffei tells us – may beenvisaged to form a giant stairway, the ascent of which affords, step-by-step, a compre-hensive knowledge of the universe. In the picture, the Count is shown taking the firststep, under Maffei’s direction (see Adams 1938: 58–9).

Unlike the solid globe, which can only be perceived as such from without, spheres –as is clear from this figure – were to be perceived from within. The global view, we mightsay, is centripetal, the spherical view centrifugal. Nor is it any accident that the percep-tion of the spheres was imaged in terms of listening rather than looking. Visual perception,insofar as it depends on the reflection of light from the outer surface of things, impliesboth the opacity and inertia of what is seen and the externality of the perceiver. Thespheres, being transparent, could not be seen, but undergoing their own autonomous rota-tions about the common centre, they could be heard: thus the motion of the spheres wassupposed to make a harmonious sound that could be registered by the sufficiently sensi-tive ear. Dating back to Pythagoras and subsequently taken up by Plato and Aristotle, thenotion of the ‘music of the spheres’ was passed on to the Middle Ages through the writ-ings of Boethius, and became integral to the ideas of the Renaissance, starting with Marsilio

123456789101234567892012345678930123456789401234567118

Dwelling• 210 •

Figure 12.2 The fourteen spheres of the world, as drawn byGiovanni Camillo Maffei of Solofra in his Scala Naturale (Venice,1564). Giovanni’s patron, the Count of Altavilla, is shown begin-ning his ascent through the spheres.

Ficino in the fifteenth century (Hallyn 1993:232). Still today, it is commonly argued thatthe space of auditory perception is sphericalin form, a sphere that surrounds (withoutenclosing) the listener at its centre. Thuswhereas we appear to be on the edge of visualspace looking in with the eye, we are alwaysat the centre of auditory space listening outwith the ear (Schafer 1985: 88, 94; cf. Ihde1976, Carpenter and McLuhan 1960). Theglobe is to the sphere, according to this argu-ment, as vision is to hearing.1

The idea of the spherical cosmos is by nomeans exclusive to the history of Europeanthought. Let me present one further example,taken from Fienup-Riordan’s (1990) accountof the lifeworld of the Yup’ik Eskimos. Hercross-sectional depiction of the cosmos asperceived by the Yup’ik, reproduced in Figure12.3, bears an uncanny resemblance toMaffei’s diagram. At the centre is the dwelling,from which roads lead in various directionsthrough the several surrounding spheres.

A person journeying far enough in any direction would eventually arrive at a pointwhere the earth folded back up into the skyland, the home of the spirits of the game. . . Not only was the earth encompassed by a canopy from above, but below its thinsurface resided the spirits of the dead, both animal and human, each in separate villages.Four or five ‘steps’ separated these two distinct but related domains.

(Fienup-Riordan 1990: 110)

Notice how in this image the surface of the earth, far from bounding the world exter-nally, is but a thin and permeable membrane dividing the world internally, between upperand lower hemispheres.

What I hope to have established, at least in outline, is that the lifeworld, imaged froman experiential centre, is spherical in form, whereas a world divorced from life, that is yet complete in itself, is imaged in the form of a globe. Thus the movement from spheri-cal to global imagery is also one in which ‘the world’, as we are taught it exists, is drawnever further from the matrix of our lived experience. It appears that the world as it reallyexists can only be witnessed by leaving it, and indeed much scientific energy and resourceshave been devoted to turning such an imaginative flight into an achieved actuality. Oneconsequence is the alleged discrepancy between what, in modern jargon, are called ‘local’and ‘global’ perspectives. Insofar as the latter, afforded to a being outside the world, is seento be both real and total, the former, afforded to beings-in-the-world (that is, ordinary peo-ple) is regarded as illusory and incomplete. Retrieving from my shelves a geology textbookpublished in 1964 – two years before the earth was first photographed from space – I readon the very first page that ‘races of men [whose] horizons are limited to a tribal territory,the confines of a mountain valley, a short stretch of the coast line, or the congested blocks

123456789101231114567892012345678930123456789401234567118

Globes and spheres • 211 •

Figure 12.3 Yup’ik cosmology in cross section.

Reproduced from Eskimo essays: Yup’ik lives and how wesee them, by A. Fienup-Riordan, published by RutgersUniversity Press, 1990, p. 111.

of a large city’ can have no conception of the true nature and extent of the world aboutthem (Putnam 1964: 3). If true knowledge is to be had by looking at the world, this state-ment is self-evidently valid. My point, however, is that this speculist assumption is preciselywhat has given us the imagery of the world as a globe. And it is this assumption, too, thatprivileges the knowledge we get from school by looking at model globes over the know-ledge we get from life by actively participating in our surroundings.

Do not misunderstand me. I am not some latter-day flat-earther or pre-Copernican. Ido not mean to deny that the earth takes the form of a globe – something that has beenknown, if not universally accepted, at least from the time of Pythagoras – or that it isone of a number of planets revolving around a rather insignificant star. My question ishow it came to pass that this globe, the planet we call Earth, was taken to be an environ-ment, or what my geology textbook called ‘the world about us’.

We can take a cue from the writings of Immanuel Kant who, in his Critique of purereason, drew a sophisticated analogy between the topological form of the earth and thatof the universe as a whole – that is, the ‘world’ conceived as the domain of all possibleobjects of knowledge. Kant first places himself in the shoes of one ignorant of the factthat the earth is global in form:

If I represent the earth as it appears to my senses, as a flat surface, with a circularhorizon, I cannot know how far it extends. But experience teaches me that wherever Imay go, I always see a space around me in which I could proceed further.

(1933: 606)

One is thus in the hapless position of realising that one’s knowledge is limited, but ofhaving no way of knowing just how limited it is. Once it is recognised, however, that theearth is a globe, and given a knowledge of its diameter, it is immediately possible to calcu-late, from first principles, its surface area. And so, even though – as we traverse the surface– new horizons are always opening up, not only can we work out, by subtraction, howmuch there remains to be discovered, but also every fresh observation can be slotted intoposition, in relation to each and every other, within a complete, unifying spatial frame-work. Thus, to obtain a comprehensive knowledge of the environment, we must alreadyhave in mind an image of the globe, or come pre-equipped with what Kant called ‘anextended concept of the whole surface of the earth’, onto which may be mapped the dataof experience (see Richards 1974: 11). Moreover the same applies to knowledge in general,which the mind sees as arrayed upon the surface of a sphere, at once continuous andlimited in extent: ‘Our reason is not like a plane indefinitely far extended, the limits ofwhich we know in a general way only; but must rather be compared to a sphere, theradius of which can be determined from the curvature of the arc of its surface . . .’ (Kant 1933: 607). In this analogy, the topology of the earth’s surface comes to stand for the fundamental idea, which the mind is said to bring to experience, of the unity,completeness and continuity of nature. Here, surely, is to be found the very essence ofthe global outlook.

Let us, then, compare an imaginary Kantian traveller, journeying across the globe insearch of new experiences to fit into his overall conception, with the Yup’ik Eskimos, inwhose cycles of everyday and seasonal movement the cosmos, as they see it (Figure 12.3),is continually being re-created (Fienup-Riordan 1990: 110–11). For both, the earth pro-vides the ground on which they move, but whereas for the Yup’ik, this movement is conducted within the world, the Kantian traveller, for whom the world is a globe, journeys

123456789101234567892012345678930123456789401234567118

Dwelling• 212 •

upon its outer surface. It is at this surface, the interface between world and mind, sensationand cognition, that all knowledge is constituted. Not only is the surface a continuous one,it also lacks any centre. Anywhere upon it can serve, in principle, equally well as a point oforigin or as a destination. Thus if the ‘world about us’ is the globe, planet earth, it is nota world within which we dwell, as is the Yup’ik world depicted with the house at its cen-tre, but one on which we dwell. The globe, of course, does have a centre, yet a journey tothe centre of the earth, as immortalised in Jules Verne’s celebrated novel, is a voyage intothe unknown, a domain of strange and terrifying primordial forces.

In short, from a global perspective, it is on the surface of the world, not at its centre,that life is lived.2 As a foundational level of ‘physical reality’, this surface is supposedalready to have been in existence long before there was any life at all. Then somehow,through a series of events of near-miraculous improbability, there appeared on it first lifeand then, very much later, consciousness. These appearances are commonly pictured interms of the addition of extra layers of being to that basic layer represented by the earth’ssurface: hence the tripartite division into lithosphere, biosphere and noosphere, correspond-ing respectively to the inorganic substance of rocks and minerals, the organic substanceof living things and the superorganic substance of human culture and society.

Although spherical imagery is employed here, the spheres are defined as layered surfacesthat successively cover over one another and the world, not as successive horizons disclosedfrom a centre. And the outer wrapping is none other than the human mind and its prod-ucts. This picture (see Figure 12.4) is the complete obverse of the medieval conception illus-trated in Figure 12.2. The difference may be considered in relation to the genesis ofmeaning. The world which the Count of Altavilla is setting out to explore in Maffei’sdiagram is itself a world of meaning which, through a kind of sensory attunement, aneducation of attention, will be gradually revealed to him as he proceeds from one level ofunderstanding to the next. This world – like the world of the Dreaming in AboriginalAustralia (see Chapter Three, p. 56) – has properties of both transparency and depth: notonly can one see into it, but also the more one looks the further one sees. By contrast, theworld depicted in Figure 12.4, insofar as it corresponds to ‘planet Earth’, consists of puresubstance, physical matter, present-ing an opaque and impenetrablesurface of literal reality upon whichform and meaning are overlain bythe human mind. That is to say,meaning does not lie in the rela-tional context of the perceiver’sinvolvement in the world, but israther inscribed upon the outer sur-face of the world by the mind ofthe perceiver. To know the world,then, is a matter not of sensoryattunement but of cognitive recon-struction. And such knowledge isacquired not by engaging directly,in a practical way, with the objectsin one’s surroundings, but ratherby learning to represent them, inthe mind, in the form of a map.

123456789101231114567892012345678930123456789401234567118


Noosphere

Biosphere

Lithosphere

Figure 12.4 Lithosphere, biosphere and noosphere.

I reserve discussion of the notions of mapping and mapmaking for the next chapter. It issufficient to note, here, the immediate connection between the apprehension of the worldas a solid globe and the idea, commonly encountered even in anthropological literature, ofthe environment as a substrate for the external imposition of arbitrary cultural form. Theworld becomes a tabula rasa for the inscription of human history.

The familiar globes of geography classrooms provide a vivid example of such inscrip-tion or covering over. Though the sea is painted blue, the continental land-masses arefrequently painted in a mosaic of contrastive colours, representing the territories of nationstates. Thus, we are led to think, has the order of human society wrapped itself around theface of the world. Yet that order, we know, has its roots in the history of colonialism, andthe attendant voyages of (principally maritime) discovery and exploration. The image of theworld as a globe is, I contend, a colonial one. It presents us with the idea of a preformedsurface waiting to be occupied, to be colonised first by living things and later by human(usually meaning Western) civilisation. Through travel and exploration, it is said, mankindhas conquered the globe. Having now filled it up, and still multiplying in numbers at analarming pace, we are urgently searching around, not just in fantasy but also in fact, fornew worlds to colonise. Not only, then, does it appear that the world existed prior to life; it also appears that life can hop from world to world and even – like a parasitic vectorflying between successive hosts – exist temporarily in worldless suspension.

The idea that the world exists prior to the forms of life that come to occupy it, andhence that each of these life-forms is itself separately encoded in a context-free vehicle, akind of free-floating capsule that can carry form from one site of occupation to another,is deeply entrenched in both biological and anthropological thought. In biology it appearsas the doctrine of genetic preformation, according to which every organism may be spec-ified, independently of the environmental context of its development, as a uniqueconfiguration of self-replicating elements (genes). Through a process of variation undernatural selection, organisms are supposed to evolve in ways that make them better adaptedto the conditions of their environments, yet the very notion of adaptation implies thatthese conditions are specifiable in advance, in terms of a set of exogenous parameters quitedistinct from the endogenous, genetically fixed parameters of the adapting organisms.There is thus one set of specifications for life, and another set for the world (see Lewontin1983). In anthropology, cultural information is made to play much the same role as isplayed by the genes in biology. Again, there is one set of specifications for the forms oflife that are carried around – as it used to be said – ‘inside people’s heads’. And there is another set for the environment, often identified with ‘nature’ or ‘the physical world’,upon which these forms are inscribed. And if we ask ‘What kind of world is this, that isan environment for every form of life yet external to all of them?’, the answer, as we haveseen, is planet Earth, the globe.

Moreover, once the world is conceived as a globe, it can become an object of appro-priation for a collective humanity. In this discourse, we do not belong to the world, neitherpartaking of its essence nor resonating to its cycles and rhythms. Rather, since our veryhumanity is seen to consist, in essence, in the transcendence of physical nature, it is theworld that belongs to us. Images of property abound. We have inherited the earth, it issaid, and so are responsible for handing it on to our successors in reasonably good condi-tion. But like the prodigal heir, we are inclined to squander this precious inheritance forthe sake of immediate gratification. Much of the current concern with the global environ-ment has to do with how we are to ‘manage’ this planet of ours. That it is ours to manage,however, remains more or less unquestioned. Such management is commonly described

123456789101234567892012345678930123456789401234567118

Dwelling• 214 •

in the language of intervention. But to intervene in the world, as we have already hadoccasion to note (Chapter Four, p. 63), implies the possibility of our choosing not to doso (Williams 1972: 154). It implies that human beings can launch their interventionsfrom a platform above the world, as though they could live on or off the environment,but are not destined to live within it. Indeed, this rendering of action towards the environ-ment as planned intervention in nature is fundamental to the Western notion of production(see Chapter Three, pp. 58–9). History itself comes to be seen as a process wherein humanproducers, through their transforming reaction on nature, have literally constructed anenvironment of their own making.

The idea is epitomised in the title of an influential volume, published in 1956, calledMan’s role in changing the face of the earth (Thomas et al. 1956). There are two pointsabout this title to which I wish to draw attention. The first is that with the world envis-aged as planet Earth, it is its face that is presented to humanity as the substrate for thelatter’s transforming interventions. This recalls my earlier observation that in the globaloutlook, life appears to be lived upon the outer surface of the world rather than from anexperiential centre within it. The world does not surround us, it lies beneath our feet.3

The second point concerns the notion of change. It is not of course the case, as wasbelieved by some of the early advocates of uniformitarianism, that the earth has persistedsince the beginning of time in homeostatic equilibrium, at least until humans came alongto upset the balance. On the contrary, it has been – and continues to be – racked bygeological forces acting on such a scale as to make the most impressive feats of humanengineering seem puny by comparison. These earth-shaping processes, however, are consid-ered to be immanent in the workings of nature. They are what the world undergoes. Butin speaking of the role of humanity, the world appears as an object of transformation.Change figures as what is done to the planet by its present owner-occupiers, human beings.It is thus exogenous rather than endogenous, not nature transforming itself, but naturetransformed through the imposition of non-natural, human design.

This is what is meant when, in ‘changing the face of the earth’, the universal agent –‘man’ – is said to have replaced the natural environment with one which is, to an ever-greater extent, artificial. Thus the construction of the human order appears to entail thedestruction of the natural one, as production entails consumption. We are, today, increas-ingly concerned to limit what are perceived to be the destructive consequences of humanactivity. My point, however, is that the very notions of destruction and damage limitation,like those of construction and control, are grounded in the discourse of intervention. Thatis to say, they presume a world already constituted, through the action of natural forces,which then becomes the object of human interest and concern. But it is not a world ofwhich humans themselves are conceived to be a part. To them, it is rather presented as aspectacle. They may observe it, reconstruct it, protect it, tamper with it or destroy it, butthey do not dwell in it. Indeed, what is perhaps most striking about the contemporarydiscourse of global environmental change is the immensity of the gulf that divides the world as it is lived and experienced by the practitioners of this discourse, and the world of which they speak under the rubric of ‘the globe’. No-one, of course, denies the serious-ness of the problems they address; there is good reason to believe, however, that many ofthese problems have their source in that very alienation of humanity from the world of which the notion of the global environment is a conspicuous expression.

This point brings me back to the distinction, mentioned earlier, between ‘local’ and‘global’ perspectives. The difference between them, I contend, is not one of hierarchicaldegree, in scale or comprehensiveness, but one of kind. In other words, the local is not

123456789101231114567892012345678930123456789401234567118


a more limited or narrowly focused apprehension than the global, it is one that rests onan altogether different mode of apprehension – one based on practical, perceptual engage-ment with components of a world that is inhabited or dwelt-in, rather than on thedetached, disinterested observation of a world that is merely occupied. In the local perspec-tive the world is a sphere, or perhaps a nesting series of spheres as portrayed in Figures12.2 and 12.3, centred on a particular place. From this experiential centre, the attentionof those who live there is drawn ever deeper into the world, in the quest for knowledgeand understanding. It is through such attentive engagement, entailed in the very processof dwelling, that the world is progressively revealed to the knowledge-seeker. Now differentcentres will, of course, afford different views, so that while there is only one global perspec-tive, indifferent to place and context, the number of possible local perspectives is potentiallyinfinite. This does not mean, however, that they are in any sense incomplete, or that theyrepresent no more than fragments of a total picture. It is only when we come to repre-sent local differences in terms of a globalising discourse that the centre from which eachperspective is taken is converted into a boundary within which every local view is seen tobe contained. The idea that the ‘little community’ remains confined within its limitedhorizons from which ‘we’ – globally conscious Westerners – have escaped results from aprivileging of the global ontology of detachment over the local ontology of engagement.

To the extent that it has been used to legitimate the disempowerment of local peoplein the management of their environments, this idea has had serious practical consequencesfor those amongst whom anthropologists have conducted their studies. To adopt a distinc-tion from Niklas Luhmann (1979), it might be argued that the dominance of the globalperspective marks the triumph of technology over cosmology. Traditional cosmology placesthe person at the centre of an ordered universe of meaningful relations, such as thatdepicted by Maffei (Figure 12.2), and enjoins an understanding of these relations as afoundation for proper conduct towards the environment. Modern technology, by contrast,places human society and its interests outside what is residually construed as the ‘phys-ical world’, and furnishes the means for the former’s control over the latter. Cosmologyprovides the guiding principles for human action within the world, technology providesthe principles for human action upon it. Thus, as cosmology gives way to technology, therelation between people and the world is turned inside out (Figure 12.1), so that whatwas a cosmos or lifeworld becomes a world – a solid globe – externally presented to life.In short, the movement from spherical to global imagery corresponds to the underminingof cosmological certainties and the growing belief in, and indeed dependence upon, thetechnological fix. It is a movement from revelation to control, and from partial know-ledge to the calculated risk.

Let me add one further comment in conclusion. I have written throughout as thoughthe characterisations of the environment, respectively, as globe and sphere were irrevo-cably opposed, and thus mutually exclusive. But this is not really so, since each viewcontains the seeds of the other. To regard the world as a sphere is at once to renderconceivable the possibility of its logical inverse, the globe; and of course vice versa. Wecould say that both perspectives are caught up in the dialectical interplay between engage-ment and detachment, between human beings’ involvement in the world and theirseparation from it, which has been a feature of the entire history of Western thought andno doubt of other traditions as well. Concretely, this is perhaps most clearly manifest inthe architectural form of the dome (Smith 1950). A sphere on the inside, a globe on theoutside, this form has a cosmic resonance of near-universal appeal. But for any society,at any period of its history, we may expect one perspective to be ascendant, and the other

123456789101234567892012345678930123456789401234567118

Dwelling• 216 •

to be associated with its more or less muted undercurrent. And my sense of the contem-porary discourse on the environment in the West is that it continues to be dominated byglobal imagery associated with the triumph of modern science and technology, but thatit is under increasing threat from those – including many anthropologists – who wouldturn to local or indigenous cosmologies of engagement for sources of insight into ourcurrent predicament.

POSTSCRIPT

Since this chapter was written, two further strands have emerged along which I think theargument can be extended. One is to relate the image of the globe, discussed here, withthat of the tree, which is currently pervasive in the representation of biodiversity. Thesecond is to show how the distinction between globe and sphere, as alternative topolo-gies of environmental awareness, crosscuts the conventional dichotomy, as it appears incontemporary environmentalist debates, between ecocentrism and anthropocentrism.

The image of living things as arrayed upon the branches of a tree will already be familiarfrom my discussion of the genealogical model in Chapter Eight (pp. 134–5). The defin-itive feature of the model, I argued, is that every creature is specified in its essential naturethrough the bestowal of attributes passed down along lines of descent, independently andin advance of its placement in the world. The idea that the world is presented to life asa surface to be occupied, at once continuous and finite in extent – in short, as the surfaceof a globe – is simply the obverse of this notion. The intrinsic connection between thegeological image of the world as a globe and the biological image of life as a tree is beau-tifully illustrated in an engraving by Johannes Christian Bendorp, dating from the turnof the nineteenth century (reproduced in Bouquet 1995: Figure 2.6). Said to depict theTree of Jesse, it shows a bay tree – on whose trunk and branches are arrayed all thedescendants of Adam and Eve – springing from a point on the surface of a solid globe.The precise location of this point is immaterial; what is significant, however, is the inscrip-tion below, which reads: ‘God created the whole family of man from one blood, to inhabitthe entire Earth’ (Bouquet 1995: 51). Thus the Earth, as a globe, is there to be colonisedby those who ‘branch out’ over it, along their several lines of descent.

Now one of the consequences of the genealogical model, as I showed in Chapter Eight(pp. 138–9), is that difference is rendered as diversity. Thus living things are classifiedand compared, and their kinds enumerated, in terms of intrinsic properties that they aredeemed to possess by virtue of genealogical connection, irrespective of their positioningin relation to one another in an environment. This is the basis for the modern conceptof biodiversity. It follows, however, that this very concept is founded in a global perspec-tive. In other words, the mode of apprehension that would reveal the totality of livingthings as a catalogue of biodiversity is also one that reveals the world as a globe in thepurview of a universal humanity. That is why the human species is itself so conspicuouslyabsent from mainstream conceptions of global biodiversity. Species can only be enumer-ated in the natural world by a humanity that has set itself above and beyond it, and that– being simultaneously everywhere and nowhere – can set the whole of nature in its sights.So far as human differences are concerned, these are typically understood in terms of aconcept of cultural diversity that is seen as analogous to biodiversity rather than as anextension of it. And the analogy, of course, serves only to reinforce the belief that what-ever differences may exist between peoples, on account of their divergent histories ofdescent, are superimposed upon a humanity that is common to all.

123456789101231114567892012345678930123456789401234567118


To pick up the second strand: contemporary discussions concerning human rights and responsibilities towards the environment, above all in global geopolitical arenas, havetended to revolve around a pivotal opposition between the positions of so-called anthro-pocentrism and ecocentrism. By anthropocentrism is usually meant an attitude which valuesall things non-human – all inanimate and animate components of the environment barringother people – solely as instrumental means to the realisation of exclusively human ends.Against this, ecocentrism is defined as that attitude which credits the world of nature –and above all, of living things in their interrelationships – with an intrinsic value quiteindependently of the purposes and activities, and even of the presence, of human beings.Yet despite (or perhaps because of ) their conventional opposition, these two positionsshare more in common than meets the eye. Both presuppose a global perspective. Forboth, ‘there is just one big environment’, identified with the order of nature (Cooper1992: 167). But by its very vastness, this all-embracing environment is profoundly aliento human experience. It is, as David Cooper puts it, ‘much too big’ to be lived in. Onecannot relate to its components. The environment we relate to, by contrast, is the onethat surrounds us, that constitutes our milieu and our ambience. And this is sphericalrather than global in its topology.4

Since we are human, the world around us must necessarily be anthropocentric: this, initself, implies no lack of participation, nor does it entail an instrumental attitude. Indeedit is decidedly odd that the term ‘anthropocentrism’ should have been adopted to denotean attitude that, more than any other, withdraws human life from active participation inthe environment. It is an attitude that might be more accurately described as ‘anthro-pocircumferentialism’. The term may be an impossibly cumbersome one; nevertheless Ibelieve we need it, if only to distinguish the discursive construction of the environmentcharacteristic of modern Western thought and science from the many pre-modern andnon-Western cosmologies that are anthropocentric in the strict sense of placing the human being at the hub of a dwelt-in world, a centre of embodied awareness that reachesout, through the activity of the senses, into its surroundings. Thus the shift from anthropocentrism to anthropocircumferentialism is tantamount to the withdrawal of thehuman presence from the centre to the periphery of the lifeworld (Figure 12.1). Andecocentrism, finally, is just the other side of the coin from anthropocircumferentialism.For once humanity is placed on the outside, surrounding the global environment, thenthe environment – now surrounded rather than surrounding – no longer holds any placefor human beings.

123456789101234567892012345678930123456789401234567118

Dwelling• 218 •

Chapter Thirteen

To journey along a way of lifeMaps, wayfinding and navigation

INTRODUCTION

Everyone has probably had the experience, at some time or other, of feeling lost, or ofnot knowing in which way to turn in order to reach a desired destination. Yet for mostof the time we know where we are, and how to get to where we want to go. Ordinarylife would be well-nigh impossible if we did not. It remains a challenge, however, toaccount for everyday skills of orientation and wayfinding. This challenge is compoundedby the considerable potential for misunderstanding surrounding the question of what itactually means to know where one is, or the way to go. For the map-using stranger, makinghis way in unfamiliar country, ‘being here’ or ‘going there’ generally entails the ability to identify one’s current or intended future position with a certain spatial or geographiclocation, defined by the intersection of particular coordinates on the map. But a personwho has grown up in a country and is conversant with its ways knows quite well wherehe is, or in what direction to go, without having to consult an artefactual map. What,then, does he have that the stranger lacks? According to a view that has found widesupport in the literatures of geography and psychology, there is no difference in principlebetween them. Both are map-users. For both, knowing where one is means identifyingone’s position in the world with a location on the map. The difference is just that thenative inhabitant’s map is held not in the hand but in the head, preserved not on paperbut in memory, in the form of a comprehensive spatial representation of his usual surround-ings. At any moment, it is supposed, he can access this mental or ‘cognitive’ map, anddetermine his location in terms of it.

In this chapter I shall argue, to the contrary, that there is no such map, and that thebelief in its existence is a consequence of the mistaken attribution to native people of asense of what it means to know one’s whereabouts that effectively treats them as strangersin their own country. Indeed the native inhabitant may be unable to specify his locationin space, in terms of any independent system of coordinates, and yet will still insist withgood cause that he knows where he is. This, as I shall show, is because places do nothave locations but histories. Bound together by the itineraries of their inhabitants, placesexist not in space but as nodes in a matrix of movement. I shall call this matrix a ‘region’.It is the knowledge of the region, and with it the ability to situate one’s current positionwithin the historical context of journeys previously made – journeys to, from and aroundplaces – that distinguishes the countryman from the stranger. Ordinary wayfinding, then,more closely resembles storytelling than map-using. To use a map is to navigate by meansof it: that is, to plot a course from one location to another in space. Wayfinding, bycontrast, is a matter of moving from one place to another in a region. But while it would

123456789101231114567892012345678930123456789401234567118


be wrong, or at least misleading, to liken the countryman’s knowledge to a map, there isa certain parallel to be drawn between the processes of knowing and mapping. Both areenvironmentally situated activities, both are carried out along paths of travel, and bothunfold over time. Just as wayfinding has to be distinguished from navigation, however,so also mapping must be distinguished from mapmaking. For the designs to which mappinggives rise – including what have been variously categorised as ‘native maps’ and ‘sketchmaps’ – are not so much representations of space as condensed histories. Thus, to putmy thesis in a nutshell, knowing is like mapping, not because knowledge is like a map,but because the products of mapping (graphic inscriptions), as those of knowing (stories),are fundamentally un-maplike. What follows is an elaboration of this argument.

COGNITIVE MAPS

At the most general level, the question of how people find their way around may be posedin terms of two alternative metaphors. Following David Rubin (1988: 375), I call thefirst a complex-structure metaphor, and the second a complex-process metaphor. Theformer, which has long been dominant in cognitive psychology, holds that even beforethe individual steps forth into the environment, he has already had copied into his mind– through some mechanism of replication – a comprehensive description of its objects,features and locations, and the relations between them. This, of course, is the cognitivemap. Having determined his current whereabouts and desired destination within the map,and having plotted the route between them, his actual movement from place to place isa perfectly straightforward, indeed almost mechanical matter of executing the prescribedcourse. Getting from A to B, in short, is explained through the harnessing of a simpleprocess, of bodily locomotion, to a complex structure, the mental map. With a complex-process metaphor, on the other hand, little or no pre-structured content is imputed tothe mind. Instead, wayfinding is understood as a skilled performance in which the trav-eller, whose powers of perception and action have been fine-tuned through previousexperience, ‘feels his way’ towards his goal, continually adjusting his movements in responseto an ongoing perceptual monitoring of his surroundings. What the first approach explainsthrough positing an isomorphism between structures in the world and structures in themind, the second explains as the unfolding of a field of relations established through theimmersion of the actor-perceiver within a given environmental context. This is theapproach favoured by ecological psychology, and it is the one I follow here.

Before pursuing an ecological approach to wayfinding, however, it is worth reflectingon the circumstances in which the notion of the cognitive map came to be introducedin the first place. At that time, some half a century ago, psychology was still in the gripof the behaviourist paradigm. Animals, including human beings, were supposed to respondmore or less automatically, in ways conditioned by previous experience, to particularenvironmental stimuli. Seeking to verify this simple model, psychologists devised numerousexperiments in which their star laboratory animal – the humble rat – was induced to runthrough a variety of mazes. Starved at the outset, having successfully negotiated the mazethe rat would be rewarded with food from a box. The idea was that through repeatedtrials, the animal would learn to take one particular path rather than another at eachsuccessive ‘choice-point’ along the route. The whole route would then be remembered asa chain of conditioned responses, such as right or left turns, triggered by the successiveappearance of particular stimuli in the form of gateways in the maze. But rats are enter-prising creatures, and they often found ways of subverting the experimenters’ intentions.

123456789101234567892012345678930123456789401234567118

Dwelling• 220 •

They would, for example, manage to climb out of the maze near the start by pushingback the cover and then run directly over the top to the food box, where they wouldclimb back down and eat. This caused some consternation in the behaviourist camp, sinceaccording to the stimulus–response model they should have had no idea of the directionin which to head off in search of food, knowing no other way than the familiar routethrough the maze, with all its twists and turns.

To further test the rats’ abilities, psychologist Edward C. Tolman and his collaboratorsdevised what they called a ‘spatial orientation’ experiment (Tolman, Ritchie and Kalish1946). A maze was first set up as shown in Figure 13.1. Starting at A, the animals hadto run across an open circular table, then through the alley CD, and finally along theroundabout route through E and F to reach the food box at G. Once they were accus-tomed to this, the original maze was replaced with the apparatus shown in Figure 13.2.Starting again at A, the animals ran across the circular table and down the alley, only tofind it blocked at one end. After returning to the table and exploring a little way downthe other radiating paths, each rat would eventually choose to run all the way out alongone of them. The overwhelming majority opted for path number 6 – the path that wouldtake them to precisely the same spot where, in the original set-up, the food box had beenlocated. This experiment seemed to provide convincing evidence that in their training forthe first maze, the rats had not merely learned a fixed sequence of steps that would leadthem reliably towards their goal. Rather, as Tolman hypothesised, they must have builtup ‘something like a field map of the environment’, upon which could be traced allpossible routes and paths and their relationships.Having located their own position and that of thefood box in terms of this map, the rats were ableto select the path, in the second maze, that leddirectly from the one to the other. In light of thisability it was clearly inadequate, Tolman reasoned,to liken the animal’s central nervous system – asthe behaviourists had done – to a telephone switch-board such that every incoming stimulus simply‘dials up’ the appropriate response. The brain wasto be compared, instead, to a ‘map control room’where stimulus-based information would becollected and collated, and where the routes wouldbe plotted that would finally determine the animal’sovert behavioural responses (Tolman 1948: 192).

Despite its provocative title, Tolman’s 1948paper – ‘Cognitive maps in rats and men’ – hadmuch to say about rats but virtually nothing abouthuman beings. Ironically, what little Tolman didhave to say about humans had nothing to do withtheir abilities of orientation and wayfinding, butwith certain psychopathologies which, he thought,could be attributed to regimes of child training thatblocked the development of properly comprehen-sive cognitive maps. Ending on a high moral tone,Tolman preached that only by inculcating the para-mount virtues of reason and tolerance could our

123456789101231114567892012345678930123456789401234567118

Maps, wayfinding and navigation • 221 •

F G

E

D

C

B

A

H

Figure 13.1 The spatial orientation experiment: theoriginal maze.

After Tolman, Ritchie and Kalish, Studies in spaciallearning I, Journal of Experimental Psychology, 36,1946.

children be furnished with maps sufficientlybroad and comprehensive to cope with ‘that greatGod-given maze which is our human world’(1948: 208). It is hard to know what the ratswould have made of this! Be that as it may, morerecent work by James and Carol Gould on thewayfinding abilities of honey bees helps to putthe rats’ capacities in perspective. For it turns outthat what rats can do, bees can do too: namely,make their way directly to a food source, alonga course never taken before. And they can dothis without involving anything that we mightdignify by terms like ‘thought’, ‘reason’ or ‘imag-ination’. The Goulds sound an appropriate noteof scepticism when they remark that the calcu-lation a bee would have to undertake in order toplan an optimal route would not be beyond asimple computer. There is no obvious reason whythe bee, or for that matter the rat, should haveany more of an understanding of the task beforeit than the computer, or why its solution shouldcall for any intelligence whatsoever (Gould andGould 1988: 224–5).

Here is what the Goulds did with their bees. First, a group of foragers were trained tofly to a feeding station in some woods out of sight of the hive. Later, individuals aboutto set off from the hive to the feeder were captured and transported, in an opaque container,to another location well off from their regular route and from which the feeder, likewise,was hidden from view. Here they were released. It was found that the bees flew straightfrom this location to the feeder, along what can only have been an entirely novel routefor them. There is no way in which they could have done this, had they been constrainedto follow a fixed sequence of steps between accustomed landmarks – as stipulated by thestimulus-response model. Instead, the Goulds suggest, the bee does what we would dounder similar circumstances: ‘she would use nearby landmarks to figure out where she is, determine in which direction her goal lies, and then depart directly towards it’ (Gould and Gould 1988: 109). She navigates, in other words, in terms of a cognitivemap. That humans do likewise was suggested by experiments conducted by Worchel (citedby Oatley 1977: 539–40), who led his subjects blindfold along two sides of a right-angledtriangle and then told them to make their way back along the hypoteneuse – a task theycompleted with considerable accuracy. The ability to update one’s position on the cogni-tive map, and thereby to keep on target despite twists and turns, is – according to KeithOatley – the basis for any kind of navigation, whether on land or at sea. But whateverthe conditions under which it is carried out, navigation ‘is a complex cognitive skill’(Oatley 1977: 537).

Comparing what the Goulds say about bees with what Oatley says about humans, wefind more than a hint of double standards. Confronted with essentially the same task, itssuccessful accomplishment by humans is attributed to complex skills whereas beesapparently do it on autopilot. I do not mean to deny that human wayfinding is a highlycomplex, skilled process. But there seems good reason to suppose that it is skilled precisely

123456789101234567892012345678930123456789401234567118

Dwelling• 222 •

Blo

ck

12

1314

15

16

17

18

1110

8

7

6

5

4

3

2

1

9

I

A

H

Figure 13.2 The spatial orientation experiment: thereplacement maze.

After Tolman, Ritchie and Kalish, Studies in spaciallearning I, Journal of Experimental Psychology, 36, 1946.

to the extent that it goes beyond the simple computational operations described by cogni-tive map theorists. For the environment within which people find their way about is not,as Tolman would have it, a ‘great God-given maze’, with all its landmarks, routes, open-ings and obstructions already laid out in advance. It is rather an immensely variegatedterrain of comings and goings, which is continually taking shape around the traveller evenas the latter’s movements contribute to its formation. To hold a course in such an environ-ment is to be attentive at all times to what is going on around you, and to respond inways that answer to your purpose. This is probably as true of rats, in their ordinaryenvironment, as it is of human beings in theirs. Rats are sensitive and intelligent crea-tures, and if their performance in experimental mazes manifests a basic computationalcapacity but no real skill, this is only because the artificial set-up in which they find them-selves is a highly impoverished one that deprives them of any opportunity for the exerciseof normal powers of discrimination and judgement.

WHAT IS A MAP ANYWAY?

The core assumption of the cognitive approach to orientation and wayfinding is, as wehave seen, that perceptually salient aspects of the structure of the world are copied intoan analogous structure in the mind (Rubin 1988: 375). This copy is said to be a map,or at least to be maplike in form. But why should this particular metaphor have beenadopted, rather than some other? Why maps rather than, say, pictures or images? Whatis the difference between a map of the world and a picture or image of the world? Anygeneral definition of a map, say Arthur Robinson and Barbara Petchenik, ‘must be basedon its being simply a representation of things in space’ (1976: 15). Yet a perspectivedrawing would satisfy this criterion, and we would surely not describe such a drawing asa map. One possible approach to defining a map, in contradistinction to the perspectivalimage, is suggested by Alfred Gell (1985). The approach rests on the idea that mapsencode beliefs or propositions about the locations of places and objects that are true (ortaken to be true) independently of where one is currently positioned in the world. Anexample of such a proposition might be that ‘Edinburgh is north of London’. One couldissue statements to this effect whether one was in London, Edinburgh, or anywhere elsefor that matter, and they would all be equally valid. In Gell’s terms, these statements –each of which is a token of the proposition in question – are non-indexical, in that theirtruth conditions are not bound to the place where they are made.

Accordingly, Gell proceeds to define the map as ‘any system of spatial knowledge and/orbeliefs which takes the form of non-token-indexical statements about the spatial locationsof places and objects’ (1985: 278–9). Now a person equipped with knowledge in thisform ought, in principle, to be able to figure out just how the world should look fromany selected point of observation. If I were hiking in the mountains, for example, I shouldbe able to state how the various peaks would appear arrayed before me, were I standingon a particular summit. Such statements, however, since they hold good only for the viewfrom that summit, and none other, are indexical of the place. Any set of beliefs and propo-sitions whose tokens are indexical in this sense, having regard for what is where for asubject positioned at a certain location, comprises what Gell calls an image (1985: 280).Thus the difference between the image and the map comes to hinge on the criterion ofthe indexicality or non-indexicality of its tokens. If our knowledge consisted only of images– that is, of token-indexical spatial propositions – then, to follow Gell’s argument, wewould never be able to hold any coherent idea about our own location in space, or about

123456789101231114567892012345678930123456789401234567118


the locations of other places relative to ourselves. We know where we are, not becausewhat we see around us matches to a certain mental image, but because this image hasitself been uniquely derived from an underlying map, at a point defined by a given setof spatial coordinates that are indifferent to our own movement. As we travel from oneplace to another, we pass through a sequence of images, each of which is specific to –and in turn permits us to identify – a particular location along the way. But the map,from which all these images are generated, remains the same wherever we are.

I shall return in due course to what Gell has to say about the nature of navigation andwayfinding. For the moment I want to focus on the implications of this way of distin-guishing between the map and the image. It is certainly true, as Gell intimates, that themere possession of a map, whether mental or artefactual, will not help you to find yourway around unless you can use it to generate location-specific images for comparison withimmediate perceptual experience. It is also true that no map will do the work that cogni-tive theorists expect of it unless the information it encodes is invariant with respect tothe location of the percipient. Consider Oatley’s assertion, for example, that the essenceof navigation lies in the ‘ability to update one’s position within the cognitive map whiletravelling’ (1977: 539). How could this possibly be done if the map keeps changing asone goes along? Oatley himself confuses the issue, when he speaks of the navigator’s cogni-tive map as ‘a process, not just a picture’ (p. 546). For if the navigator is to look to themap for directions, it can be neither process nor picture, neither embodying his ownmovement nor representing any particular scenes along the route. ‘We only update maps’,as Gell observes, ‘when the geography of the world changes, not whenever we move aboutourselves’ (1985: 274). Ultimately, the justification for extending the map metaphor intothe domain of cognition must lie in the assumption, more often than not unstated, thatwhat the map affords is a representation of things in space that is independent of anyparticular point of view.

This assumption, however, raises problems of its own. One of the difficulties that cartog-raphers often face in their attempts to explain the nature of maps is that the very fields,of cognition and communication, from which they might find appropriate analogues havealready seized upon the map as an analogue from cartography. ‘When non-cartographicwriters use the term “map”’, as Robinson and Petchenik say, ‘they seem to mean that itis possible to take isolated incidents, experiences, and so on, and arrange them intellec-tually so that there is some coherence, some total relation, instead of individual isolation’(1976: 4). Thus scientists refer to their theories as maps, into which can be fitted the dataof observation, while anthropologists are inclined to attribute a similar maplike quality toculture and society (for example, Leach 1976: 51), on the grounds that it furnishes anoverarching framework of concepts and categories for the organisation of otherwise frag-mentary sensory experience. These, and many other similar metaphorical usages make itappear natural and self-evident that actual maps should function in the same way, asschematic representations of the real world, which do not index any position but uponwhich it should be possible to plot the position of everything in relation to everythingelse. Now most people in Western societies, educated since their schooldays in theconventions of modern cartography, probably do tend to think of maps as representationsof this kind. But whether the artefacts and inscriptions that have at one time or anotherbeen designated as maps actually satisfy the requirement of non-indexicality, is moot. Thequestion, in short, is: are maps maplike?

David Turnbull, arguing from the perspective of a sociologist of science, makes acompelling case to the effect that they are not. The idea that maps are independent of

123456789101234567892012345678930123456789401234567118

Dwelling• 224 •

any point of view, that the propositions they encode are equally valid wherever one standsin the world, is, Turnbull contends, a myth – though it is one that has been avidly culti-vated in the name of science and objectivity (Turnbull 1989: 15). The reality is that nomap, however ‘modern’ or sophisticated the techniques of its production, can be whollydivorced from the practices, interests and understandings of its makers and users. Or toput it another way, every map is necessarily embedded in a ‘form of life’. And to theextent that it is so embedded, it must fail on the criterion of non-indexicality. As Turnbullexplains, ‘all maps are in some measure indexical, because no map, representation or theorycan be independent of a form of life’ (1989: 20). At first glance, this argument seems torun directly counter to Gell’s insistence that a representation can only be a map insofaras the propositions encoded therein are non-indexical. Closer examination, however, revealsa certain slippage in the meaning of indexicality. Is indexing a place the same as indexinga form of life? If the map discloses a perspective or ‘point of view’, is this a view in theworld, as it appears from a particular place, or a view of the world, filtered through theconcepts, categories and schemata of a received cultural tradition? Could a map be non-indexical in the first sense and indexical in the second?

Consider an example to which both Gell and Turnbull refer. Micronesian mariners,1

who are used to voyaging across hundreds of miles of open sea between often tiny islands,know the bearing of any island from any other by its so-called ‘star course’ – that is, bya list of stars whose successive rising or setting points, during the night, indicate the direc-tion in question. The expert mariner has committed to memory an entire compendiumof star courses, each unique to a particular pair of islands, and it is in this compendium,according to Gell, that his ‘map’ consists. Now it is clearly the case that any statementof the course between one island A, and another island B, will not depend for its validityon one’s current position at sea. Thus star courses ‘have the essential map property ofnon-token-indexicality; they do not change truth value according to where they are uttered’(Gell 1985: 284). Yet it is also fair to say, with Turnbull, that the principles upon whichthe Micronesian mariner’s map is constructed are securely embedded within the perceptsand practices of traditional seafaring, and therefore that it requires a knowledge of thiscultural context to be able to ‘read’ and understand the map. It would appear, in short,that while the map indexes a tradition, it is non-indexical with regard to location. Thesame, moreover, could be said of ‘modern’ maps, constructed on scientific principles withthe aid of sophisticated technological gadgetry. Modern science and technology, asTurnbull remarks (1991: 36), are as dependent on tradition for their successful trans-mission as is Micronesian seafaring lore. And no more than Micronesian maps can modernmaps be understood without taking into account ‘the world view, cognitive schema orthe culture of the mapmaker’ (Turnbull 1989: 20).

There is, however, something deeply paradoxical about this argument. For to separatetradition from locality, or culture from place, is also to divorce traditional knowledge fromthe contexts of its production in the environmentally situated experience of practitioners.Thus the form of life is reduced to a ‘world view’ or ‘cognitive schema’ – a set of rulesand representations for the organisation of sensory experience that individuals carry intheir heads and that are available for transmission independently of their bodily activityin the world. It is as though culture were received along lines of traditional transmissionfrom ancestors, and imported into the sites of its practical application. But this is to fall right back into the classical view of culture as a map, the analogy – as Bourdieu (1977:2) points out – ‘which occurs to an outsider who has to find his way around in a foreignlandscape and who compensates for his lack of practical mastery, the prerogative of the

123456789101231114567892012345678930123456789401234567118


native, by the use of a model of all possible routes’. So here is the paradox: actual mapsare made to appear indexical with regard to cultural tradition only by a rendering ofculture as non-indexical with regard to locality. The placing of maps within their culturalcontext is paralleled by the displacing of culture from its context in the lifeworld. How,then, are we to resolve this dilemma? How can we hold on to the commonsense notionthat maps retain a certain invariance as we move about, that they do not continuallyrecompose themselves to reflect the particularities of wherever we happen to be, while yetrecognising their embeddedness in locally situated practices? My answer, in brief, will bethat what maps index is movement, that the vision they embody is not local but regional,but that the ambition of modern cartography has been to convert this regional vision intoa global one, as though it issued from a point of view above and beyond the world.

HOW TO SEE THE WORLD FROM EVERYWHERE AT ONCE

When you stand at a particular spot, everything appears from a certain angle, while muchof the environment will likely be hidden from view behind prominent foreground features.Stand at another spot, and things will appear differently. In order to have any concep-tion of the overall configuration of one’s environment, it would seem necessary to be inpossession of some kind of totalising scheme into which every one of these location-specificperceptual images could be integrated. This, as we have seen, is an argument commonlyadduced to justify positing the existence of cognitive maps. It is an argument, however,that assumes a snapshot theory of vision, as if one could only ever see, in perspective,from a fixed point of observation. ‘Is not to see’, as Merleau-Ponty asks rhetorically, ‘alwaysto see from somewhere?’ He proceeds to answer, however, in the negative (Merleau-Ponty1962: 67). To take up his own example, the house next door may be viewed from thisside or that, from inside or outside, or even from up above if one were to fly overhead.But what I see is none of these appearances; it is the house itself, in all its concrete actu-ality. The form of the house is progressively disclosed to me as I move around and about,and in and out, not as the sum of a very large number of images, arrayed in memorylike frames on a reel of film, but as the envelope of a continually changing perspectivalstructure. Observation, Merleau-Ponty claims, consists not in having a fixed point of viewon the object, but ‘in varying the point of view while keeping the object fixed’ (1962:91). Thus the house is not seen from somewhere but from nowhere – or rather fromeverywhere (pp. 67–9).

In keeping with his ecological approach to visual perception, James Gibson presents anargument along very similar lines. Animals and people, Gibson writes, see as they move,not just in the intervals between movements. Such ambulatory vision takes place along whathe calls a ‘path of observation’. A path is to be understood not as an infinite series of discrete points, occupied at successive instants, but as a continuous itinerary of movement.Thus the environment one sees is neither ‘seen-at-this-moment’ nor ‘seen-from-this-point’.On the contrary, ‘what one perceives is an environment that surrounds one, that is every-where equally clear, that is in-the-round or solid, and that is all-of-a-piece’ (Gibson 1979:195–7). But if the features of this environment are revealed as one travels along paths ofview, rather than projected from a sequence of points of view, where do these paths begin,and where do they end? And if we see not at this moment in time, but over a certain period,how long is this period? Such questions cannot be precisely answered. Of a minor featurewe might say, after only cursory exploration, that we have seen it all. But of a complex,varied and extensive terrain, although we may have criss-crossed it along innumerable paths,

123456789101234567892012345678930123456789401234567118

Dwelling• 226 •

we may still feel there is more to be discovered. As for our perception of the environmentas a whole, what else can this be than the outcome of a lifetime’s observation, along all thepaths we have ever taken? This is what Gibson means when he asserts that perceiving theworld over a sufficient length of time, and along a sufficiently extended set of paths, istantamount to perceiving it ‘as if one could be everywhere at once’ (p. 197).

It is critically important to distinguish this sense of omnipresence from that impliedby the conventional notion of the ‘bird’s-eye view’ (Gibson 1979: 198–9). The latter, ofcourse, has nothing to do with the way birds in flight actually see, but rather describeshow we imagine the world would look from a point of observation so far above the earth’ssurface that the entire territory with which we are familiar from journeys made at groundlevel could be taken in at a glance. The higher one goes, it is supposed, the more one’s vision transcends the locational constraints and narrow horizons of the view fromthe ground. And by the same token, the more apparently maplike it becomes. Robinsonand Petchenik are right to point out that the analogy between the map and the bird’seye view is potentially misleading, not only because of their different geometries of projec-tion, but also because the map is ‘a construction, an abstraction, an arrangement ofmarkings that relates to spatial “reality” only by agreement, not by sensory testability’(1976: 53). Nevertheless, anyone who has flown over familiar country by plane will havebeen astonished, on the one hand, by how strange it looks, and on the other, by howclosely the view from the window resembles a topographic map of the same territory.There is nothing strange, however, about the environment perceived from everywhere, inthe sense adduced by Merleau-Ponty and Gibson, nor do you have to leave the groundto perceive it in this way. It is not a view from ‘up there’ rather than ‘down here’, butone taken along the multiple paths that make up a country, and along which people comeand go in the practical conduct of life. Our perception of the environment as a whole,in short, is forged not in the ascent from a myopic, local perspective to a panoptic, globalone, but in the passage from place to place, and in histories of movement and changinghorizons along the way.

The same point could be made, following Edward Casey (1996: 30), through a contrastbetween vertical and lateral modes of integration. In the vertical mode, embraced bymodern cartography as well as by cognitive map theorists, local particulars obtained byobservation on the ground are fitted within an abstract conception of space so as to forma representation of the world as though one were looking down upon it from ‘up above’.While the eyes of the body remain close to the ground, the mind’s eye – which is witnessto this maplike representation – is up with the birds. The lateral mode of integration, by contrast, presupposes no such division between mind and body. For the work of inte-gration is performed by the organism as a whole as it moves around, purposefully andattentively, from place to place. Such movements do not merely connect places that arealready located in terms of an independent framework of spatial coordinates. Rather, theybring these places into being as nodes within a wider network of coming and going. Caseyrefers to this network of interplace movement as a region – that is, ‘an area concatenatedby peregrinations between the places it connects’ (1996: 24). Evidently, when Gibsonspeaks of perceiving the environment from everywhere at once, that ‘everywhere’ is neitherspace, nor a portion of space, but a region in this sense. Likewise, every ‘somewhere’ isnot a location in space but a position on a path of movement, one of the matrix of pathscomprising the region as a whole. In short, whereas everywhere-as-space is the world asit is imagined from a point of view above and beyond, everywhere-as-region is the worldas it is experienced by an inhabitant journeying from place to place along a way of life.

123456789101231114567892012345678930123456789401234567118


This idea of the region may be illustrated by means of three ethnographic examples.Among the Walbiri, an Aboriginal people of western central Australia, the entire countryis perceived ‘in terms of networks of places linked by paths’ (Munn 1973a: 215). Originallylaid down through the movements of ancestral beings in that formative era known as theDreaming, these paths are continually retraced in the journeys of the living people whotake after them. As they relate the stories of these journeys, Walbiri men and women maydraw web-like figures in the sand whose basic components are lines and circles. Every lineconveys a journey to or from camp, while every circle conveys the act of making campby walking all around it. Rather similarly for the Ongees, a group of hunter-gatherersinhabiting the island of Little Andaman in the Bay of Bengal, places are brought intobeing at the confluences of the paths of movement of humans, animals and spirits. Askedby the ethnographer, Vishvajit Pandya, to draw the places where humans and spirits live,Ongee informants responded by sketching lines of movement (straight for humans, wavyfor spirits), leading to the demarcation of the various places at their intersections.2

The world of the Ongees, Pandya concludes, ‘is not a preconstituted stage on which thingshappen, but rather an area or region created and constructed by the ongoing practice ofmovement’ (Pandya 1990: 777). My third example is taken from A. Irving Hallowell’sstudy of the Saulteaux (Ojibwa), hunters and trappers of the Berens River district nearLake Winnipeg in Canada. In Saulteaux experience, to move in a certain direction isalways to travel from place to place. This is so not only for human persons, but also forthe sun, the moon and the winds, all of which are held to be persons of a kind. Thus‘what we refer to abstractly as cardinal directions are to them the homes of the winds, theplaces they come from. Similarly, east is thought of as the place where the sun rises; west,the place where it sets; south is the place to which the souls of the dead travel, and theplace from which the summer birds come’ (Hallowell 1955: 191). For the Saulteaux, then,as indeed for the Ongee and the Walbiri, ‘everywhere’ is not a space but a region concaten-ated by the place-to-place movements of humans, animals, spirits, winds, celestial bodies,and so on.

KNOWING AS YOU GO

We can now return to the paradox I introduced earlier. If our knowledge of the environ-ment is embedded in locally situated practices, how come that it retains a certain constancyas we move about? If all knowledge is context-dependent, how can people take their know-ledge with them from one context to another? For clues towards a resolution I turn onceagain to the work of David Turnbull. One of Turnbull’s aims is to break down theconventional distinction between so-called indigenous knowledge and Western science. Hedoes so by emphasising that all knowledge, of whatever kind and historical provenance,is generated within a ‘field of practices’ (1989: 61). And since practices must be carriedout by particular people in particular places, all knowledge – including that which we callscience – must be inherently local. Let me set aside for the time being the contrary thesis,which Turnbull confusingly appears to entertain at the same time, that the context forboth indigenous and scientific knowledge is something like a worldview or cognitiveschema, by nature detached from the local sites of its practical expression. I have alreadydrawn attention to the dangers of falling back on a concept of culture that divorces know-ledge and its transmission from environmentally situated experience. My present concernis with another difficulty in Turnbull’s argument. For while on the one hand, he insiststhat a common characteristic of all knowledge systems is their ‘localness’, he also argues,

123456789101234567892012345678930123456789401234567118

Dwelling• 228 •

on the other, that what is critical to the growth and reproduction of any knowledge systemis the work that goes into moving its diverse components – including practitioners, theirknow-how and skills, technical devices and standards of evaluation – from one local siteof knowledge production to another (Turnbull 1993a: 30).

Consider the case of Western science. According to what might be called the ‘official’view of science, data recorded by means of standardised procedures in diverse locations arefitted into a framework of theory consisting of propositions that are strictly non-indexicalwith regard to place. What happens in practice, however, is a good deal more messy. Notonly is it unclear where data collection ends and theory building begins, but also there isno unified body of theory under which all of experience can be subsumed. Rather, thereare as many theoretical growth-points as there are sites of practical investigation, and thecharacter of each is conditioned by circumstances peculiar to each place. Much of the labourof science, Turnbull argues, lies in attempts to establish the connectivity and equivalencethat would render procedures developed and results obtained in one local context applica-ble in another (1993a: 37). But if science calls for the constant movement of personnel,knowledge and techniques from place to place, and the assemblage, in each, of inputs ofheterogeneous provenance, how can it also share the characteristic of localness? As a systemof knowledge, science cannot be rooted in any particular place or places, but must ratheremerge from the total network of interplace relations constituting its field of practice.Furthermore, if that is so for science, then it should be equally so for any other knowledgesystem. As Turnbull himself puts it, ‘all knowing is like travelling, like a journey betweenthe parts of a matrix’ (1991: 35). So what is this matrix? It is, of course, a region in thesense defined above – that is, as the sum of journeys made.

My point is that knowing, like the perception of the environment in general, proceedsalong paths of observation. One can no more know in places than travel in them. Rather,knowledge is regional: it is to be cultivated by moving along paths that lead around,towards or away from places, from or to places elsewhere. Conceived as the ensemble ofsuch place-to-place movements, the notion of region, far from denoting a level of gener-alisation intermediate between local particulars and global universals, offers a way out ofthis kind of dichotomous and hierarchical thinking. As every place, through the move-ments that give rise to it, enfolds its relations to all others, to be somewhere is to beeverywhere at once. Rephrased in our terms, what Turnbull proposes is a compelling argu-ment to the effect that all knowledge systems, including science, are integrated laterallyrather than vertically. The philosopher Joseph Rouse makes much the same point inarguing that ‘we go from one local knowledge to another rather than from universal theo-ries to their particular instantiations’ (Rouse 1987: 72). In light of the foregoingconsiderations, I would prefer to say that we know as we go, from place to place. Thisdoes not, however, alter the basic point, which is that science is distinguished from othersystems of knowledge by the lengths to which it goes to present itself as if it were verti-cally integrated, as if the scientist’s task were to fit data to theory rather than to put theknowledge that has brought him to one place to work in setting off towards another. Tocreate this illusion, science has to suppress, or to hide from view, the social labour involvedin establishing equivalences and connections across places (Turnbull 1996: 62). In this,moreover, it is aided and abetted by modern cartography, which has been similarlyconcerned to establish its scientific credentials through its claim to produce accurate andobjective representations of a world ‘out there’.

Cartographers, like scientists, and indeed like practitioners of any other knowledgesystem, draw their material from all manner of sources, through both direct observation

123456789101231114567892012345678930123456789401234567118


and inquiry into local tradition. The collection and collation of this material may takethem – or agents operating on their behalf – on innumerable and often lengthy journeys.None of this, however, appears in the final form of the modern, ‘scientific’ map. To thecontrary, one of the most striking characteristics of the modern map is its elimination, orerasure, of the practices and itineraries that contributed to its production (Turnbull 1996:62). In the words of Michel de Certeau, ‘the map, a totalising stage on which elementsof diverse origin are brought together to form a tableau of a “state” of geographical know-ledge, pushes away into its prehistory or into its posterity, as if into the wings, theoperations of which it is the result or the necessary condition’ (1984: 121). Just as science,in the official view, is charged with the task of integrating site-specific data into an over-arching, unified framework of theory, so the mission of cartography is ostensibly one ofrepresenting the ‘geographic facts’ on the ground within a single, universal system of spatialcoordinates (Edney 1993: 55). The ideal is a perfect congruence between the world andits representation, and progress is measured by the degree of approximation towards it.Thus in the work of the modern cartographer, knowledge generated through movementfrom place to place within a region is presented as if it issued from a totalising visionabove and beyond the world. In short, cartography transforms everywhere-as-region, theworld as experienced by a mobile inhabitant, into everywhere-as-space, the imaginary‘bird’s-eye view’ of a transcendent consciousness.

The same transformation, of course, is worked on the ordinary perception of the environ-ment by the theory of cognitive maps. As in the modern artefactual map, so too in its‘mental’ analogue, all those movements of coming and going through which people developa knowledge of their environment are pushed into the wings, to recall de Certeau’s phrase,leaving the map as a fait accompli, final and complete, the product of a process of makingthat begins with the layout of the world and ends with that layout copied into the mind.Any journeys undertaken beyond that point are supposed to belong to the phase of map-using rather than mapmaking, and therefore to play no further part in the formation ofthe map itself. The traditional Micronesian seafarer, in this view, is just as much a map-user as is the modern marine navigator with his charts and compass, even though his skill‘is entirely mental and perceptual, using no instruments of any kind’ (Oatley 1977: 537).But whereas modern artefactual maps have their authors, designers or manufacturers, theorigins of traditional mental maps appear lost in the mists of time. Indeed to say of suchmaps that they are ‘traditional’ is virtually tantamount to an admission that they have nomaker or makers, but rather that they ‘make themselves’ – or that like myths, followingLévi-Strauss’s celebrated aphorism, they ‘think themselves out’ through the medium ofmen’s minds and without their knowledge (Lévi-Strauss 1966a: 56). In any case theassumption is that the map is made before it is used, that it already exists as a structurein the mind, handed down as part of a received tradition, prior to the traveller’s venturingforth into the world.

My contention, to the contrary, is that people’s knowledge of the environment under-goes continuous formation in the very course of their moving about in it. To return toa distinction which I introduced at the outset, this is to account for such knowledge in terms of the generative potentials of a complex process rather than the replication ofa complex structure. That process consists in the engagement of the mobile actor-perceiverwith his or her environment. As I have already suggested, we know as we go, not beforewe go. Such ambulatory knowing – or knowledgeable ambulating – cannot beaccommodated within the terms of the conventional dichotomy between mapmaking andmap-using. The traveller or storyteller who knows as he goes is neither making a map

123456789101234567892012345678930123456789401234567118

Dwelling• 230 •

nor using one. He is, quite simply, mapping. And theforms or patterns that arise from this mapping process,whether in the imagination or materialised as artefacts,are but stepping stones along the way, punctuating theprocess rather than initiating it or bringing it to a close.My perspective, in short, accords with what RobertRundstrom has called ‘process cartography’, in whichmapping is seen as ‘open-ended, ongoing, alwaysleading to the next instance of mapping, the next map’(Rundstrom 1993: 21). In what follows, I first showin more detail how mapping differs from mapmaking.I then turn to the distinction between mapping andmap-using. All wayfinding, I argue, is mapping; allnavigation map-using. Thus mapping is to map-usingas wayfinding to navigation. The overall structure ofthe argument is summarised in Figure 13.3.

MAPPING IS NOT MAPMAKING

‘Mapping’ and ‘mapmaking’, according to Denis Wood, ‘do not mean the same thing’(1992: 32). The difference, in his view, is akin to that between speaking and writing.Wood thinks of mapping as a capacity universal to humans, established along with othercapacities of the human mind-brain through a process of evolution under natural selec-tion. But the fact that all human beings are capable of mapping does not mean that theyall make maps. Likewise, just because all humans can speak does not mean they all write.Whereas mapping, like speaking, might be regarded as a ‘universal expression of indi-vidual existence’, mapmaking, like writing, has to be seen as ‘an unusual function ofspecifiable social circumstances arising only within certain social structures’ (Wood 1993a:50). In other words, the emergence of mapmaking belongs not to the evolution ofhumanity but to its history. Yet the difference between mapping and mapmaking, just asthat between speaking and writing, is for Wood a very fine one. It is not the differencebetween outwardly expressing an idea and ‘capturing’ that expression in an alternativemedium. For one thing, mapping is no more the externalisation of a map that alreadyexists in the mapper’s head than is speaking the externalisation of a thought. Rather, bothmapping and speaking are genres of performance that draw their meanings from thecommunicative contexts of their enactment. It follows, for another thing, that neithermapmaking nor writing can serve to transcribe pre-existent thoughts or mental represen-tations onto paper. The map, like the written word, is not, in the first place, thetranscription of anything, but rather an inscription. Thus mapping gives way to mapmakingat the point, not where mental imagery yields an external representation, but where theperformative gesture becomes an inscriptive practice (Wood 1993a: 53).

Wood illustrates his argument with a nice example. Two boys have been playingrollerblade hockey. At home over dinner, one explains the layout of the court by gesturingwith his hands and fingers over a place mat. The other does the same at school, to impressa friend, but in this case (it is during an art class) he gestures with pencil in hand, overa sheet of paper. Whereas nothing remains of the first boy’s gestures on the mat, thoseof the second leave a trace in the form of an inscription, a sketch-map, that can bepreserved and reproduced indefinitely beyond the context of its production. We may

123456789101231114567892012345678930123456789401234567118


Mapping =Wayfinding

Mapmaking =Cartography

Map-using =Navigation

versus

Figure 13.3 The relations between mapping,mapmaking and map-using: a summary.

suppose that the two boys were of equal ability, and moreover that the first would havehad ready access to pencil and paper had he needed it. So why did the second make amap and the first not? The answer, for Wood, lies in the nature of the communicativesituation. In general, just as much as in this exemplary instance, it is the situation – atonce social and political – that calls for the map. And while the difference betweengesturing with an inscribing tool and gesturing without might seem slight, the socio-political consequences are immense. It is the ‘fine line of . . . inscription’, Wood concludes,‘that differentiates . . . mapping . . . from mapmaking, and mapping societies frommapmaking societies, in the latter of which it is the inscriptive property of the artefactualmap that permits it to serve the interests of the power elites who control the mapmakingprocess (as well as those who would contest them)’ (1993a: 53).

Now while I agree with Wood that there is an important distinction to be made betweenmapping and mapmaking, I would draw it along different lines. Before doing so, however,we need to be more precise about the meaning of mapping. Wood himself seems unableto make up his mind whether the term refers to a cognitive capacity, to actual movementin the environment, or to the narrative reenactment of journeys made. At one point hetells us that mapping ‘is the way we humans make and deploy mental maps’ (1992: 32),while at another he dismisses the concept of the mental map only to declare that mapping‘is really just . . . getting around’ (1993a: 53). Yet in his example of the two boys, mappingappears to consist neither in having a pre-existent ‘map in the head’, nor in bodily move-ment on the ground, but in a kind of retrospective storytelling. It seems to me that thenotion of an evolved capacity for mental mapping is deeply flawed. One could hardlyexpect any such capacity to spring, fully formed, from an individual’s genetic make-up,in advance of his or her entry into the lifeworld. It would rather have to undergo devel-opment in the very unfolding of the individual’s life within an environment. Thus thelife-historical process of ‘getting around’ – or in a word, wayfinding – would appear tobe a condition for the emergence of a ‘mapping capacity’, rather than a consequence ofits application. This leaves us with the third sense of mapping – the retelling of journeysmade (or possibly the rehearsal for journeys to be made) – as perhaps the most appro-priate. I admit, however, that the distinction between wayfinding and mapping is not hardand fast. For one way of retelling the story of a journey is to retrace one’s steps, or thesteps of ancestors who made the journey in the past. In effect, since travelling from oneplace to another means remembering the way, all wayfinding is mapping, though not allmapping is wayfinding. I return to this point below.

For the time being, let us continue to regard mapping as the re-enactment, in narrativegesture, of the experience of moving from place to place within a region. In this sense, bothboys in Wood’s example were engaged in mapping. The fact that one left no trace whereasthe other produced a lasting inscription has no appreciable bearing on the nature of theactivity as such. The sketch-map that emerged, as the trace of the second boy’s gestures,was a more or less incidental by-product of the mapping process, not its ultimate goal.Rundstrom makes much the same point in his account of mapping among Inuit of thecentral and eastern Canadian Arctic. An Inuit traveller, returning from a trip, could recountevery detail of the environment encountered along the way, miming with his hands theforms of specific land and sea features. Such gestural performance, after a long journey,could last many hours. It could also, given appropriate tools and materials, generate aninscription. Many of these inscriptions were produced at the instigation of Western explor-ers who made contact with the Inuit. They were often astonished at the accuracy of whatthey regarded as ‘native maps’. But for Inuit mappers it was the performance that mattered

123456789101234567892012345678930123456789401234567118

Dwelling• 232 •

– ‘the recapitulation of environmental features’ – rather than any material artefacts orinscriptions to which it gave rise (Rundstrom 1990: 165). Undoubtedly the vast majorityof maps that have ever been produced in human societies, like those of the Inuit, have beenimprovised on the spot within a particular dialogic or storytelling context, and without anyintention for their preservation or use beyond that context. This applies, for example, tothe web-like sand drawings of the Walbiri, to which I have already referred (Munn 1973b:196). ‘Most maps for most of the time’, as Wood observes, ‘have probably been ephemeral,scratched in sand or snow, or, if committed to a more permanent medium, immediatelycrunched up and thrown away’ (1993b: 83, see Lewis 1993: 99).

In the course of producing such a map, the mapper takes his interlocutors on a tourof the country, and as he does so his moving hand, which may or may not hold aninscribing implement, traces out the paths taken and the sights or landmarks encounteredalong the way. Of the maps produced in aboriginal times by the Saulteaux, Hallowellnotes that ‘their purpose was not to delineate a section of the country as such, but toindicate a route to be followed, and the emphasis was upon a succession of landmarksroughly indicated in their relations to one another’ (Hallowell 1955: 195). Malcolm Lewis’sstudies of native North American and Inuit maps have shown that they invariably rest ondeictic principles: that is, they point to things, revealing aspects of how they look as oneproceeds along a path of observation from ‘here’ to ‘there’ (Lewis 1993: 102). Even incontemporary Western societies, whose inhabitants are bombarded on a daily basis withimages founded upon cartographic geometries of plane projection – where they live, asWood puts it, ‘map-immersed in the world’ (1992: 34) – people continue to describetheir environment, to themselves and others, by retracing the paths of movement theycustomarily take through it rather than by assigning each of its features to a fixed loca-tion in space. ‘When we are asked for directions’, as Barbara Belyea notes, ‘few of us canresist pointing and waving our arms, or tracing the traveller’s route over the surface ofhis map. The gesture becomes a part of the map, a feature of its reception’ (Belyea 1996:11, my emphasis). It may be misleading, Belyea suggests, to liken the inscriptive processto writing, as though the purpose of the exercise were to represent the features of thelandscape in the same way that writing is supposed to represent the spoken word. Forthe graphs on the map are not representations of anything. Every line is rather the traceof a gesture, which itself retraces an actual movement in the world. To read the map istherefore to follow the trace as one would the path of the hand that made it.3

The analogy between mapping and writing, however, may be closer than Belyea thinks.For much of its history, at least in the Western world, writing was understood not as therepresentation of speech but as a means by which what has been said or told could becommitted to memory (Carruthers 1990). Throughout the Middle Ages, as David Olsonnotes, ‘written records were thought of and treated as reminders rather than representa-tions’ (Olson 1994: 180). And the same was true of medieval maps, which served asmemoranda of itineraries, providing directions and advice to the traveller who wouldundertake the same journey (de Certeau 1984: 120). In the history of writing as in thatof mapping, remembering gradually gave way to representation over the same period –from the fifteenth to the seventeenth century – that also saw the rise of modern scien-tific discourse. De Certeau has shown how, in the course of this transition, the map ‘slowlydisengaged itself from the itineraries that were the conditions of its possibility’. For sometime, maps would continue to be illustrated with pictures of ships, landforms, people andbeasts of various descriptions, winds and currents, and the like. Subsequently dismissedas quaint decorations, these figures were really fragments of stories, telling of the journeys,

123456789101231114567892012345678930123456789401234567118


and the incidents that took place along them, from which the map resulted. But even-tually, the map won out over these pictorial figurations, eliminating all remaining tracesof the practices that produced it (de Certeau 1984: 120–1). Thus the making of mapscame to be divorced from the experience of bodily movement in the world.4 The carto-grapher has no need to travel, indeed he may have no experience whatever of the territoryhe so painstakingly seeks to represent. His task is rather to assemble, off-site, the infor-mation provided to him – already shorn of the particular circumstances of its collection– into a comprehensive spatial representation. It is of course no accident that preciselythe same task is assigned, by cognitive map theorists, to the mind in operating upon thedata of sense.

It is at the point where maps cease to be generated as by-products of story-telling, andare created instead as end-products of projects of spatial representation, that I draw the line between mapping and mapmaking. In effect, mapmaking suppresses, or ‘bracketsout’, both the movements of people as they come and go between places (wayfinding),and the re-enactment of those movements in inscriptive gesture (mapping). It therebycreates the appearance that the structure of the map springs directly from the structureof the world, as though the mapmaker served merely to mediate a transcription from oneto the other. I call this the cartographic illusion (see Figure 13.4). One aspect of this illu-sion lies in the assumption that the structure of the world, and so also that of the mapwhich purports to represent it, is fixed without regard to the movement of its inhabitants.Like a theatrical stage from which all the actors have mysteriously disappeared, the world– as it is represented in the map – appears deserted, devoid of life. No-one is there;nothing is going on. Suppose, for example, that I describe a journey I have made bytracing a path with my finger over the surface of a topographic map. Once the map has

been folded and put away, nothingof this would remain. So far as themap’s representation of the world isconcerned, I may as well have nevermade the trip. Had I, alternatively,traced my path with a pencil, theresulting lines would be deemed tohave added nothing to the map, butrather to have defaced it. To restorethe map, they would have to berubbed out! Either way, my gesturedoes not become part of the map butis excluded from it, as is my originalmovement from the world it repre-sents.5 This is in marked contrast tothe maps of native North AmericanIndians and Inuit, as described bysuch scholars as Lewis, Rundstromand Belyea, which actually grow, lineby line, with every additionalgesture. So do the charts used byMicronesian seafarers, which ‘liter-ally get larger, coconut-palm rib bycowrie shell, and stick by stone’

123456789101234567892012345678930123456789401234567118

Dwelling• 234 •

Mapmaking

SituatedMovement

World Map

Wayfinding Mapping

Figure 13.4 The cartographic illusion. The environmen-tally situated movement entailed in both wayfinding andits narrative re-enactment (mapping) is bracketed out tocreate the illusion that the form of the map arises, inmapmaking, as a direct transcription of the layout of theworld.

(Wood 1992: 31). And so, too, do our own sketch-maps. In these instances the devel-opment of the map, as a ‘pattern of interconnected lines’ (Belyea 1996: 6), parallels thatof the region, as a network of coming and going. But the modern topographic map doesnot grow or develop, it is made. And just as the process of its production is eliminatedfrom the final form of the product, so the world it describes is not a world in the making,but one ready-made for life to occupy.

It is this, finally, that lies behind the distinction between the map and the picture, asalternative descriptions of the same country. For those of us schooled in the conventionsof modern cartography, the distinction may seem obvious enough. Maps are supposed tofurnish an objective record of the disposition of things in space, that is strictly indepen-dent of any point of view, whereas pictures show how these things might be experiencedby a subject positioned somewhere in that space, or moving through it (Turnbull 1989:15). It is widely believed, as Svetlana Alpers observes, that ‘maps give us the measure ofa place and the relationship between places, quantifiable data, while landscape pictures areevocative, and aim rather to give us some quality of a place or the viewer’s sense of it.One is closer to science, the other is art’. Anything on the map that evokes the experi-ence of place or movement is dismissed by the scientific cartographer as ‘mere decoration’;anything in the picture that conveys factual information about spatial location is dismissedby the artist as ‘mere topography’ (Alpers 1983: 124–6). But for the Dutch painters anddraughtsmen of the seventeenth century, who are the subjects of Alpers’s study, theseboundaries between maps and pictures, and between science and art, would have made little sense. Mapping and picturing were, for them, one and the same, having astheir common aim ‘to capture on a surface a great range of knowledge and informationabout the world’ (1983: 122). As mapmaking triumphed over mapping, however, and ascartographers sought to dissociate themselves professionally from artists, so maps werestripped of their pictorial attributes. Thus historians of cartography, viewing the develop-ment of mapmaking in retrospect, are able to present it as having progressed from beingan ‘art’ to being a ‘science’, replacing subjective fancy with hardwon and independentlyverifiable factual information (Edney 1993: 56). Art, in the words of Brian Harley, wasgradually ‘edged off the map’ (Harley 1989: 4). But to edge art off the map is also toedge human actor-perceivers off the world, to push their direct, sensory experience intothe wings, and to consign their narratives of movement and travel to the realms of fable,fantasy and hallucination.

WAYFINDING IS NOT NAVIGATION

‘Navigation’, writes Edwin Hutchins, ‘is a collection of techniques for answering a smallnumber of questions, perhaps the most central of which is “Where am I?” ’ (Hutchins1995: 12). So – to return to a question I raised at the outset – what does it mean toknow where one is? What would one need to know in order to feel that the question hasbeen satisfactorily answered? First of all, according to Hutchins, one must possess somerepresentation of space – a map – whether internal or external, inscribed in the mind oron a sheet of paper, within which every object or feature in one’s environment is assigneda determinate location. One has then to be able to establish a coherent set of correspon-dences between what is depicted on the map and what is visible in one’s surroundings.From these it should be possible to identify one’s current position in the world with aspecific location on the map. Only then has the question of where one is been answered(Hutchins 1995: 12–13). Alfred Gell, in an article to which I have already referred, argues

123456789101231114567892012345678930123456789401234567118


along much the same lines. To know where one is, in Gell’s view, it is not enough tohave formed a perceptual image of the environment as seen from some place. This imagehas to be matched to that generated from the map (mental or artefactual) at a particularspatial location. ‘Navigation’, according to Gell, ‘consists of a cyclic process whereby imagesgenerated from maps are matched up against perceptual information, and perceptual imagesare identified with equivalent coordinates on a map’ (1985: 280). This process of matchingis essentially the same as what Hutchins means by ‘establishing correspondences’, such as,for example, when we say ‘this here’ (pointing to contours on the map) corresponds to‘that there’ (pointing to the outline of a hill on the horizon).

Now while Gell takes as his principal ethnographic example the classic case ofMicronesian seafaring, Hutchins chose to study the practices of nautical navigation onboard a large modern naval vessel. Both writers insist, however, that reduced to its bareessentials, navigation is a cognitive task that all of us face all the time as we find our wayabout, whether at sea or on land. Navigational techniques may of course be distinguished,as Gell admits, both in terms of their complexity and the volume of information handled,and in terms of the extent to which this information is published or transmitted by rotememorisation. But none of this, he claims, alters the fact that ‘the essential logical processesinvolved in all way-finding, from the most elementary and subliminal, to the most complexand laborious, are identical’ (Gell 1985: 286). For Hutchins, likewise, we are all naviga-tors in our everyday lives, as the following passage reveals:

When the navigator is satisfied that he has arrived at a coherent set of correspondences,he might look to the chart and say ‘Ah, yes; I am here, off this point of land.’ And itis in this sense that most of us feel we know where we are. We feel that we have achievedreconciliation between the features we see in our world and a representation of thatworld.

(1995: 13, my emphasis)

Yet as soon as Hutchins takes us on board ship, and introduces us to the work of thenavigators on the bridge, things look rather different. For it turns out that establishingcorrespondences between features on the chart and features in the environment is extremelydifficult, and calls for specialised skills that can only be acquired through lengthy trainingand hands-on experience. To reconcile the chart with the territory, as Hutchins explains,one has to imagine how the world would look from a point of view – that of the ‘bird’seye’ – from which it is never actually seen, save from an aircraft or satellite. The ordi-nary passenger, untutored in the techniques of navigation, is quite unable to do this, andmay confess to being baffled by maps and charts. He cannot, in other words, translatefrom his on-board experience of motion as ‘moving through a surrounding space’ to thedepiction of motion on the chart as ‘that of an object moving across a space’. Navigators,on the other hand, become so used to thinking of the movement of the ship from thispeculiar perspective – as if they were manoeuvring it about like a counter on a game-board – that they find it difficult to imagine this movement, any more, from the ordinarypassenger’s perspective (Hutchins 1995: 62).

I intend to argue, in accord with Hutchins’s ethnography but contrary to his generalclaim, that we are no more navigators in our everyday lives – in finding our way around ina familiar environment – than we are cartographers when we retrace these movements in narrative. Navigation (or map-using) is, I contend, as strange to the ordinary practices ofwayfinding as is cartography (or mapmaking) to ordinary practices of mapping. It would be

123456789101234567892012345678930123456789401234567118

Dwelling• 236 •

hard to imagine why we should find the navigator’s charts so baffling, or why his skillsshould be so specialised, if they were but analogues of cognitive structures and capacitiesthat we use all the time. Thus Gell, along with others who have had resort to the notionof cognitive maps, is surely wrong to regard wayfinding and navigation as processes of a similar or even identical kind. For when we move about, we do not normally think of ourselves as piloting our bodies across the surface of the earth, as the navigator pilots hisship across the ocean. Nor do we have to think in this way in order to know, at any moment,where we are. This is because the question ‘Where am I?’ is not ordinarily answered in termsof a location in space, determined by the intersection of an independent set of coordinates.Hutchins to the contrary, it is not in this sense that most of us feel we know where we are.Indeed I may know precisely where I am and yet have no idea of my geographic location.For it is not by assigning the position where I currently stand to certain spatial coordinatesthat an answer to the ‘where’ question is arrived at, but rather by situating that positionwithin the matrix of movement constitutive of a region.

To amplify this point, let me compare two, admittedly fictional, scenarios. In the firstyou are walking with a friend through unfamiliar terrain, equipped with a topographicmap. Arriving at a place that affords a good panoramic view, your friend stops to ask,‘Where are we?’ You look around, pointing to various landmarks which you proceed tocorrelate with markings on the map. Finally, indicating with a finger a particular spot onthe map’s paper surface, you declare ‘We are here’. In the second scenario, you are walkingin familiar country around your home, with a companion who is a stranger to the area.Once again, on arrival at a certain place, your companion puts the same question, ‘Whereare we?’ You may respond in the first instance with a place-name. But then, realising thatthe name alone leaves him none the wiser, you might go on to tell a story about the place– about your own association with it, about other people who have lived and visited there,and about the things that happened to them. Now in the second case you have no needto consult an artefactual map, nor would it be of any avail to you, not because you haveresort instead to a map inside your head, but because knowing your present whereaboutshas nothing to do with fixing your location in space. As someone who has lived in acountry, and is used to its ways, knowing where you are lies not in the establishment ofa point-to-point correspondence between the world and its representation, but in theremembering of journeys previously made, and that brought you to the place along thesame or different paths. In the first scenario, of course, you have no knowledge of thiskind. Having never visited the country before you do not know where you are, in thesense you do when on home ground, even though you may be able to locate your ownposition, and that of everything else, with pin-point accuracy on your map.

For those who know a country, in short, the answers to such basic questions as ‘Wheream I?’ and ‘Which way should I go?’ are found in narratives of past movement. It is inthis respect, as noted earlier, that wayfinding and mapping become one and the same: tofollow a path is also to retrace one’s steps, or the steps of one’s predecessors. And in thisrespect, too, wayfinding differs fundamentally from navigation, just as mapping differsfrom map-using. For when navigating in a strange country by means of a topographicmap, the relation between one’s position on the ground and one’s location in space, asdefined by particular map coordinates, is strictly synchronic, and divorced from any narra-tive context. It is possible to specify where one is – one’s current location – without regardto where one has been, or where one is going. In ordinary wayfinding, by contrast, everyplace holds within it memories of previous arrivals and departures, as well as expectationsof how one may reach it, or reach other places from it. Thus do places enfold the passage

123456789101231114567892012345678930123456789401234567118


of time: they are neither of the past, present or future but all three rolled into one.Endlessly generated through the comings and goings of their inhabitants, they figure notas locations in space but as specific vortices in a current of movement, of innumerablejourneys actually made. Taking this view of place as my starting point, I now want toshow how wayfinding might be understood not as following a course from one spatiallocation to another, but as a movement in time, more akin to playing music or story-telling than to reading a map.

PATHS, FLOWS AND THE PASSAGE OF TIME

The inspiration for this move comes from Gibson, and follows from his insight – whichI explored in an earlier section – that the environment is perceived not from multiplepoints of view but along a path of observation. Rejecting both of the dominant psycho-logical approaches to wayfinding, as chains of conditioned responses to environmentalstimuli and as navigation by means of cognitive maps, Gibson proposes an alternative,‘the theory of reversible occlusion’ (1979: 198). In brief, the theory states that one knowsthe way in terms of the specific order in which the surfaces of the environment come intoor pass out of sight as one proceeds along a path. Suppose, for example, that you arewalking along a street in town, or through a valley in the countryside. The surfaces youcan see – the facades of buildings in the one case, or the ground rising on either side inthe other – comprise a vista. As Gibson explains, a vista is ‘a semienclosure, a set ofunhidden surfaces, . . . what is seen from here, with the proviso that “here” is not a pointbut an extended region’. But now, as you turn the corner into another street, or reachthe brow of the ridge at the head of the valley, a new set of surfaces, previously hidden,looms into view, while those of the original vista disappear from sight. The passage fromone vista to another, during which the former is gradually occluded while the latter opensup, constitutes a transition. Thus to travel from place to place involves the opening upand closing off of vistas, in a particular order, through a continuous series of reversibletransitions. It is through this ordering of vistas, Gibson maintains, that the structure ofthe environment is progressively disclosed to the moving observer, such that he or shecan eventually perceive it from everywhere at once (Gibson 1979: 198–9).

Gibson’s notion of wayfinding through reversible occlusion has been further developedin recent work by psychologist Harry Heft (1996). We have already seen how the forms of environmental features are revealed as the envelopes of a continually modulating pers-pective structure along a path of observation. Now this flow of perspective structure, as Heft points out, also specifies the observer’s own movements relative to the layout of theenvironment. As every path of travel gives rise to its own distinctive flow pattern, so everysuch pattern uniquely specifies a certain path. To find one’s way, Heft argues, means totravel along a particular route so as to generate or recreate the flow of perspective structurepeculiar to the path leading to one’s destination (1996: 122). One remembers the route asa succession of vistas connected by transitions, rather as one might remember a piece ofmusic as a series of thematic sections linked by bridge passages. Just as with musicalperformance, wayfinding has an essentially temporal character (1996: 112): the path, likethe musical melody, unfolds over time rather than across space. In this connection, it isimportant to remind ourselves of Gibson’s contention that every path should be conceivedas a unitary movement, and not as a potentially infinite set of adjacent points (Gibson 1979:197). In music, a melodic phrase is not just a sequence of discrete tones; what counts is therising or falling of pitch that gives shape to the phrase as a whole. Likewise in wayfinding,

123456789101234567892012345678930123456789401234567118

Dwelling• 238 •

the path is specified not as a sequence of point-indexical images, but as the coming-into-sight and passing-out-of-sight of variously contoured and textured surfaces.

In this respect, too, the theory of wayfinding advanced here differs profoundly fromthat which Gell has caricatured under the rubric of ‘mapless practical mastery’, and whichhe attributes, inter alia, to Bourdieu (Gell 1985; see Bourdieu 1977: 2). ‘We can suppose’,writes Gell, ‘that practical mastery of the environment consists of possessing completeknowledge of what the environment looks like from all practically-available points of view’.The master traveller, equipped with such knowledge, remembers the journey from A toB as a ‘chain of linked landscape images’, each particular to a certain point along theroute, selected from the total stock of images filed in memory. As he proceeds on his wayhe will pause, every so often, to check that what he sees from the spot where he standscorresponds to the image he has on file (Gell 1985: 274–5). Our argument, to the contrary,is that mastery consists in knowing what the environment looks like from all practicallyavailable paths of view, that what the traveller remembers are vistas and transitions ratherthan location-specific images, and that keeping track is a matter of regenerating the flowof perspective structure over time. Now for Gell the theory of mapless practical mastery,taken on its own, could not possibly work, since it would leave the traveller bereft of anymeans to formulate navigational decisions. It is all very well to know that you are currentlywhere you ought to be – that what you see around you matches your expectations for acertain stage in your journey. But this alone will not tell you in which direction to go toreach the next checkpoint. Nor, if what you see does not match any of the images in thechain for the particular journey you are making, do you have any way of working outhow to get back on track. In short, to go from A to B, or from any point to any otheralong the way, you need to be able to ascertain their relative locations in space. And this,Gell reasons, requires a map.

If it were true that all wayfinding consisted of navigation between fixed points, Gell’sargument would be unassailable. But it is not. Ordinary movement in a familiar environ-ment lacks the stop-go character of navigation, in which every physical or bodily manoeuvre(displacement in space) is preceded by a mental or calculative one (fixing the course).‘Finding one’s way’ is not a computational operation carried out prior to departure froma place, but is tantamount to one’s own movement through the world. To recapitulatemy earlier point, we know as we go, not before we go. Thus the operation is not completeuntil one has reached one’s final destination: only then can the traveller truly claim tohave found his way. The notion of ‘finding’ has here to be understood in its originalsense of exploratory movement, at once improvisatory and assured, guided by past expe-rience and by a continual monitoring of fluctuations not only in the pattern of reflectedlight but also in the sounds and ‘feel’ of the environment. There is no better illustrationof this than the example that Gell himself uses in an attempt to prove, to the contrary,that wayfinding is based on the execution of pre-formulated ‘navigational decisions’ (1985:282). This is the case of Micronesian seafaring. In a classic paper on the subject, ThomasGladwin describes how, at every moment during a voyage, the mariner is attentive to ‘a combination of motion, sound, feel of the wind, wave patterns, star relationships, etc.’,all of which – through comparison with remembered observations from past experience– translates into ‘a slight increase or decrease in pressure on the steering paddle, or agrunted instruction to slack off the sail a trifle’ (Gladwin 1964: 171–2). Quite unlike theEuropean navigator, with his charts and compass, the Micronesian seafarer feels his waytowards his destination by continually adjusting his movements in relation to the flow ofwaves, wind, current and stars.6 In this respect his activity does not differ in principle

123456789101231114567892012345678930123456789401234567118


from that of the terrestrial traveller who responds to the flow of perspective structure ashe journeys through a landscape. Both are essentially engaged in projects of wayfindingrather than navigation: thus Hallowell’s observation that for the Saulteaux, direction alwayshas the meaning of ‘toward such-and-such a place’, is paralleled by Gladwin’s that theMicronesian mariner proceeds as if he were constantly within sight of land (Hallowell1955: 190–1, Gladwin 1964: 173). And once it is recognised that the wayfinder’s multi-sensory monitoring is of flows, not images, and that flows specify paths and not spatiallocations, Gell’s objections to the idea of mapless practical mastery fall away.

Micronesian seafaring resembles terrestrial wayfinding in one other critical respect: everyjourney is apprehended and remembered as a movement through time rather than acrossspace. Islands, for the mariner, are not pinned down to specific spatial or geographic loca-tions, nor does he imagine his craft to be covering the distance over a planar surface fromone such location to another. Throughout the voyage he remains, apparently stationary,at the centre of a world that stretches around as far as the horizon, with the great domeof the heavens above. But as the journey proceeds the island of embarkation slips everfarther astern while the destination island draws ever closer. At the same time an islandoff to one side, selected as a point of reference for the voyage, is supposed to swing pastthe boat, falling as it does so under the rising or setting positions of a series of stars. Thefact that the reference island (etak) is normally invisible below the horizon, and may noteven exist at all, has been a source of puzzlement to many interpreters who – assumingthat the mariner’s task is to navigate from one spatial location to another – have proposedthat the etak is used to obtain a locational fix. Nothing in what the mariners themselveshave to say, however, suggests that it serves any such purpose. The alleged bearing of theetak does not enter into any numerical computation. Rather, pointing to the etak is themariner’s way of indicating where he is in terms of the temporal unfolding of the voyageas a whole (Hutchins 1995: 87–8). We have already seen how, in terrestrial wayfinding,a route from one place to another is remembered as a temporally ordered sequence ofvistas. In much the same way, the Micronesian mariner remembers an inter-island voyageas a sequence of etak segments, each of which begins as the reference island falls underone particular star and ends as it falls under the next in line. At any movement, themariner will know what segment he is in. As it swings beneath the horizon, from segmentto segment, the etak island marks in its movement the passage of time, just as do the sun,moon and stars overhead, in theirs. Completion of the penultimate segment should bringthe mariner, at length, to the final ‘etak of sighting’, as the island for which he is boundhoves into view.

THE WORLD HAS NO SURFACE

One further contrast remains to be drawn between wayfinding and navigation, and it takesus back to the cartographic notion of the map as a representation of some portion of theearth’s surface. The following ‘official’ definition of the map, issued by the InternationalCartographic Association, is exemplary:

A map is a representation normally to scale and on a flat medium, of a selection ofmaterial or abstract features on, or in relation to, the surface of the Earth or of a celestialbody.

(cited in Robinson and Petchenik 1976: 17)

123456789101234567892012345678930123456789401234567118

Dwelling• 240 •

Now the idea that the world is presented to the traveller as a surface to be traversedpresupposes the specialised, ‘bird’s-eye view’ of the cartographer or navigator. Indeed theworld can only be perceived to have an exterior surface by a mind that is situated aboveand beyond it. In ordinary wayfinding however, whether on land or at sea, the world is apprehended from within. One makes one’s way through it, not over or across it. Ofcourse the traveller encounters surfaces of diverse kinds – of solid ground, water, vegeta-tion, buildings, and so on – and it is largely thanks to the responses of these surfaces tolight, sound and the pressure of touch that he perceives the environment in the way he does. For the mariner the ocean, with its subtle differences of tint and colour, sculptedby the wind into waves and ripples, and breaking up around the boat into foam and spray, presents an infinitely variegated and ever changing surface. Likewise for thepedestrian, making his way along a forest track, the surface of the ground is a patchworkof mud, furrowed by the imprint of previous journeys, puddles, fallen leaves, brokenboughs, and outcropping rocks and stones. These are surfaces, however, in the world, notof the world. That is to say, they are formed on the interface, not between matter andmind, but between solid or liquid substance and the gaseous medium (air) in whichhumans live and breathe, and which affords movement and sensory perception.7 In shortfor its manifold inhabitants, journeying along their respective ways of life, the world itselfhas no surface.

I noted earlier the parallel between the tracing of paths on the ground in wayfindingand the tracing of lines on paper (or in sand, snow, etc.) in mapping: indeed to the extentthat all wayfinding is mapping, these are one and the same. Our conclusion, however,that for the mapper or wayfinder the world has no surface, calls for some qualification ofthe view, for which I argued above, that mapping is an inscriptive process. This need notbe so. If a map consists of a network of interconnected lines, each corresponding to apath of movement through the world, there is no necessary reason why these lines shouldbe inscribed on a surface. One could think of the gesturing hand, in mapping, as a weavinghand rather than a drawing hand, and of the result as something more akin to a cat’scradle than a graph. The lines of the map could be threads, wires or sticks. Micronesianmariners used coconut leaf ribs to map the intersecting courses of ocean swells (Turnbull1991: 24). Or to take a familiar example from a contemporary urban context, one couldconstruct a route map for the London Underground out of stiff wire, soldered at theintersections, and it would serve just as well as the conventional printed versions. Thefact that the map is generally reproduced on paper is a matter of obvious practical conven-ience, but not of logical necessity. The meaning of the map lies entirely in its routes andintersections, whereas the paper surface has no significance whatsoever. To read the mapis to trace a continuous path from one station to another, without regard to their respec-tive locations on the surface. With the modern topographic map it is quite otherwise, forin this case the paper surface of the map stands for nothing less than the surface of the earth. One of the most revealing indicators of this change in the significance of the map-surface, corresponding to the transition from mapping to mapmaking, lies in the appearance of frame boundaries. Native maps, as Belyea points out (1996: 6), arenever framed. A line or border drawn around and enclosing such a map would have no meaning. The frame of the topographic map, by contrast, defines the portion of theearth’s surface that the map purports to represent. Thus the appearance of borders around the map corresponds to the disappearance of the itineraries and practices that giverise to it.

123456789101231114567892012345678930123456789401234567118


CONCLUSION

There is a paradox at the heart of modern cartography. The more it aims to furnish aprecise and comprehensive representation of reality, the less true to life this representa-tion appears. ‘To present a useful and truthful picture’, as Mark Monmonier writes, ‘anaccurate map must tell white lies’ (Monmonier 1991: 1). But the reason for the discrep-ancy between truth and accuracy is not quite what Monmonier claims it to be. It is notthat the map must leave things out if critical information is not to be drowned in a welterof ever finer particulars. It is rather that the world of our experience is a world suspendedin movement, that is continually coming into being as we – through our own movement– contribute to its formation. In the cartographic world, by contrast, all is still and silent.There is neither sunlight nor moonlight; there are no variations of light or shade, noclouds, no shadows or reflections. The wind does not blow, neither disturbing the treesnor whipping water into waves. No birds fly in the sky, or sing in the woods; forests andpastures are devoid of animal life; houses and streets are empty of people and traffic. Todismiss all this – to suggest that what is excluded in the cartographic reduction amounts,in Monmonier’s words, to a ‘fog of detail’ – is perverse, to say the least (Wood 1992:76). For it is no less than the stuff of life itself. Were one magically transported into thelooking-glass world behind the map, one would indeed feel lost and disoriented, as in afog. But the fogginess is a function not of the amount or density of detail but of thearrestation of movement. Detached from the flow of which each is but a moment, detailssettle like an opaque precipitate upon the surface of the earth. Little wonder, then, thatthe cartographer feels the need to sweep them up, or that the navigator prefers to brushthem aside in plotting a course!

The ordinary wayfinder, on the other hand, is not generally troubled by detail. Quiteto the contrary, the richer and more varied the texture of the environment, the easier itis to find one’s way about. But above all, wayfinding depends upon the attunement ofthe traveller’s movements in response to the movements, in his or her surroundings, of other people, animals, the wind, celestial bodies, and so on. Where nothing movesthere is nothing to which one can respond: at such times – as before a storm, or duringan eclipse – the experienced traveller can lose his bearings even in familiar terrain. Theseobservations should finally lay to rest the cartographic illusion, namely that the world ispre-prepared as a stage upon which living things propel themselves about, from one loca-tion to another. Life, in this view, is an internal property of objects, transported uponthe exterior surface of a lifeless earth. In the view I have set forth here, by contrast, theworld is not ready-made for life to occupy. Contrary to the assumptions of cartographersand cognitive map theorists, life is not contained within things, nor is it transported about.It is rather laid down along paths of movement, of action and perception. Every livingbeing, accordingly, grows and reaches out into the environment along the sum of its paths.To find one’s way is to advance along a line of growth, in a world which is never quitethe same from one moment to the next, and whose future configuration can never befully known. Ways of life are not therefore determined in advance, as routes to be followed,but have continually to be worked out anew. And these ways, far from being inscribedupon the surface of an inanimate world, are the very threads from which the living worldis woven.

123456789101234567892012345678930123456789401234567118

Dwelling• 242 •

Dwelling - Pedro P. Ferreira · the sphere, conjuring up a transparent lifeworld which is perceived...

Documents

Transcript of Dwelling - Pedro P. Ferreira · the sphere, conjuring up a transparent lifeworld which is perceived...